I am analyzing some bearing vibration data from an induction motor that drives a centrifugal compressor. The unit is 2500 HP, 3600 rpm, has split sleeve bearings with oil fed lubrication and a shim-pack style coupling. Sorry that I do not have the actual spectra to present, but here is the mix of data (velocity,ips) I have:

ODE Axial: 1x=0.05, 2x=0.07
ODE Vert: 1x=0.015, 2x=0.02
ODE Horz: 1x=0.03, 2x=0.075

DE Axial: 1x=0.015, 2x=0.045
DE Vert: 1x=0.12, 2x=0.04
DE Horz: 1x=0.1, 2x=0.15

3x, 4x, 5x exist, but all are much lower and in decending amplitude. Coupling-end compressor bearing vibration has a dominant 1X vibration in all directions, but peak amplitudes are low--about 0.05 maximum.

Is this misalignment, looseness, shaft bow, some combination thereof, or something totally different?

Is the disparity with a dominant 1x in the ODE Vert with a dominant 2x and strong 1x in the ODE Horz indicative of anything in particular? Also, what does the higher axial vibration at the ODE mean, considering all the DE vibration is in the other directions?

Thanks to everyone who chimes in.

"Is the disparity with a dominant 1x in the ODE Vert with a dominant 2x and strong 1x in the ODE Horz indicative of anything in particular?"

should have read...

Is the disparity with a dominant 1x in the DE Vert with a dominant 2x and strong 1x in the DE Horz indicative of anything in particular?

Sorry about the DE-ODE confusion.

I had a situation where there was an increase
in vertical readings that was caused by a plugged supply line.This was on a feedwater pump
but could possibly happen with plugged air filters on a compressor.
If this is a new situation and unusual it might be something to look at.

My first gut feeling would be an alignment issue, but not a serious one, based on the amplitudes. Shim packs are pesky things to align. One thing I've found over the years is that they can build up stresses in the coupling if the are out of alignment, and the shim pack is tightened up. I usually laser align, and loosen the shim pack once alignment has been reached, then tighten it back up.

quote:

One thing I've found over the years is that they can build up stresses in the coupling if the are out of alignment, and the shim pack is tightened up. I usually laser align, and loosen the shim pack once alignment has been reached, then tighten it back up.

I don't understand what is gained by that procedure. Can you explain a little more what the purpose of that procedure would be. Thx.

Presence of multiple 1x harmonics is normal for plain bearings. The amplitudes do not seem to be high and the only minor concerning factor is 2x slighly dominating the spectra. Trending them over time will help to assign severity.

To rule out a bowed shaft situation, reading phase at 12,3,6,9 o'clock will be helpful. For misalignment test, phase readings accross the coupling (including axial) will be the best indicator.

Probably orbit on a bearing will give you same information as far as misalignement/looseness is concerned (although I am not sure about usefulness of the orbit obtained using seismic accelerometers on the housing).

Close examination of the TWF is also a must. Slight frequency modulation or randomness could be detected in a TWF thus pointing to the source of harmonics.

Regards,
David

Pete, if two coupling halves with a steelflex gric are misaligned, as they coupling halves are brought into alignment, the "grid" is free to move. In a shim pack coupling, the shims and coupling halves are connected by bolts. The bolts hold the shims together, which doesn't allow the shims to move.

So, if after final alignment, you loosen the bolts in the shim pack, and allow them to "find their place", then tighten them back up, you don't induce stresses into the shim pack.

I agree with Pete. What does this gain?