Hi,

we had a problem on 500 kw motor, did vibration and it showed high axial reading indication misalignment. We did alignment and it was ok, equipment worked fine. But when we did reading vertical vibration are normal but horizontal are over allowed limits on DE drive and NDE drive.

We have basic equipment, measuring only overall vibration and acceleration...

What would indicating faults if vibration are high in vertical or horizontal direction? I am wondering if you could help me to learn to diagnose faults based on reading on horizontal/vertical, article ore something...? I am aware that I would need FFT equipment but I do not... Thanks

I'd make a survey starting at the bearing centerline and work down to the floor.If there is motion of the baseplate at the floor I'd tighten the anchor bolts and check again.


Does your equipment have an output jack?
If it accesses the vibration signal, not a processed overall value signal (4-20ma), I'd try to process it with one of the free FFT progams for PCs. Some have limited ability for resolution down.

http://audacity.sourceforge.net/

searching for gram5, spectrogram, and free fft gets tons of hits

I have not tried this one yet
http://www.plant-maintenance.com/freestuff/570.shtml

Thanks, but I can not use FFT I have instrument that measuring overall vibration. What faults indicate horizontal or vertical vibration?

I know that axial indicate misalignment...

Enrique, I would suggest refering to Ralph's recent post about onsite training. I assume it must be very discouraging if all you have is overall 4-20 monitoring. But asking for help with such little info is like saying that my car has a shake in the front end. It could be a loose lugnut or the poor skunk my kid ran over last night. Perhaps you can search the site for radial, vertical & axial.

That is what I have, I wish I have better... I just would like to try to learn what possible fault would be it I have vertical or horizontal increased vibration levels, or if is the same level...thanks

I agree with the above comments about the importance of getting a spectrum. At this point we don't know whether your vibration is 1x, 1x+harmonics, 2*LF, or something else.

But in general, if you have a significant difference between H and V readings, it may indicate resonant amplification.

You might carefully try temporary bracing to see if it changes the vibration level to investigate resonance.

Also, if perhaps this might be 2*LF, you might try a soft foot check. For example carefully loosen 1 foot at a time while running and see if the vibration goes down. Foot related 2*LF vibration seems more common on 2-pole motors.

Your hand-held meter: does it do various frequency ranges? Yes. Take a reading at 200 or 500 Hz, one at 1000 Hz and another at 2 kHz. And does your unit allow integration? 3 values in accel and 3 in IPS can help reduce to an area of interest.

You can take this appraoch to narrow some things down a little. If an A-F bearing has a high number in the low range and its level at the high range isn't much change, then you know you have more of a fundament problem (none of this is set in stone nor takes away from FFT). Next you would need to eliminate some fundamental problems such as power off - vibration disappears. Vary operational parameters if you can. All of which are more involved and more difficult with the least chance of success. Ever take a 2X4 and load a shaft? Stiffen a structure and frame! Play the funky chicken - inquiring minds want to know.

Make your case and justify more tools/equipment/etc.... Not much but you can't provide much.

What kind of electrical connector is provided to this motor - is it rigid conduit, or flexible conduit, or something else?

It was a good point about switching ranges and units as an aide to guessing the frequency.

IF we make a BIG ASSUMPTION that all of the vibration comes from one frequency, then we can guesstimate the frequency by comparing the readings in various units:

f = 3670 * A / V
f = 19100 * V / D
where
f = frequency in cpm
A = acceleration in g's pk/0
V = velocity in inches / sec pk/0
D is displacement in mils (0.001") pk/pk

again, note this depends on an ASSUMPTION which may or may not be correct.