Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,607
Reply with quote  #1 
I have a motor on a 3-stage Centac compressor, around 500 HP.  Pretty much every problem I have ever diagnosed on these motors (22 units spanning 15 years) has been using the MOA position which has a good solid flange the bearing is seated in.  For the last few years I have been using a triaxial sensor at this location to capture Ax, H, V.  

I have one motor that was noticeably "noisier" in early October, noisy enough that the mechanics noticed it and said something, and they rarely notice anything.  When I say "noisy", it's not audible, and really doesn't show up as discrete frequencies in the spectrum as a bad bearing normally does.  It's really just the way it "feels"... feels like it has fine gravel in it.  I was expecting to see discrete frequencies, but just really don't.  The broadband noise at lower frequencies has been growing the last several months.  

PeakVue is showing 120 Hz with harmonics, which is not unusual for these motors when they are first installed... we usually have to adjust the rear "support" bolts to eliminate this after installation of a new motor.  These are C-face motors with minimal outboard support.  This motor showed 120 Hz when first installed, but none since (in the standard spectrum).

Normally, when I have vibration that I can "feel" but not hear or see in the spectrum, I usually think it's electrical, especially if it's on a drive.  I am leaning towards a rotor bar issue since this is a constant-speed induction motor.  How often do you see a "bad" bearing showing just low amplitude broadband vibration?  Lubrication is not an issue as I lubricate these motors myself, based on noise in a 10kHz spectrum.

And I know I can check rotor bar issues by taking a current spectrum using an amp clamp, but that is such a giant pain in the butt these days, that I haven't done that it years, nor am I likely to.  I could just tell them to change the motor out, but they've been on a run of bad luck with motors (motor shop issues mainly) in this compressor room, that I hate to make a call unless I'm fairly sure it's necessary.

Your thoughts? AC Rotor Bar PeakVue.png 
AC Rotor Bar.png


__________________
"The trend is your friend"
Noknroll

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 750
Reply with quote  #2 
Rusty
What are the frequencies I've circled? looks like 1/2 order harmonics and/or side bands?theirs more that I didn't circle.
Edit:
1/2 X, shaft fretting? just thinking out loud
AC Rotor Bar PeakVue.png

RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,607
Reply with quote  #3 
It's a 6315 bearing, and none of those peaks are bearing defects, nor are they sidebands of any of the major peaks. Do peaks that small in a PV spectrum for a 3600 rpm machine really indicate anything?

Normally if I have a bearing loose on a shaft I see well-defined sidebands spaced at the difference of the shaft and the bearing inner race turning speed.  People sometimes say (or think) the bearing is "spinning" on the shaft, but actually it just creeps along at a speed a little slower than the shaft, often around 50 rpm slower.  Or I will see classic stair-stepped harmonics of 1X - but there will be a significant 1X because once you get clearance between the shaft and inner race, there will be imbalance.

There really isn't anything definitive, except the widening noise floor.  Other than that, or what I can "feel" with my hand, there is nothing to indicate a problem, imo.

__________________
"The trend is your friend"
fburgos

Sr. Member
Registered:
Posts: 463
Reply with quote  #4 
This one makes me remember one severe case of dust contamination... Tomorrow I'll get the data and pictures

Noknroll

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 750
Reply with quote  #5 
all good Rusty, just a stab in the dark and I wouldn't be concerned with that spectrum either. I just threw that in because you seemed to be looking for something.
OLi

Sr. Member
Registered:
Posts: 1,519
Reply with quote  #6 
Someone fiddled with the VFD? just guessing to check a alternative, they do misbehave sometimes.
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,607
Reply with quote  #7 
No VFD. If it was on a drive I’d definitely call it as a drive issue. Can’t recall ever seeing this symptom on a non-VFD or non-DC unit. Anyone still monitor a DC drive machine?
__________________
"The trend is your friend"
ivibr8

Sr. Member
Registered:
Posts: 133
Reply with quote  #8 
Some more thinking out loud......
Can you obtain data while shutting machine down?  Perhaps providing clue if it is electrical in nature?

Jim P
electricpete

Sr. Member
Registered:
Posts: 528
Reply with quote  #9 
It beats me. I'd be inclined to look at a higher-resolution velocity spectrum.  Maybe what looks like an increased noise floor in your low-res velocity waterfall will turn into a multitude of identifiable peaks such as pole pass sidebands or bearing defect frequencies (or maybe not). 

Also look for pole pass side bands around running speed harmonics log scale.
HuskerTim

Avatar / Picture

Member
Registered:
Posts: 44
Reply with quote  #10 
Rusty

Does your motor look like this 600HP 2P D-flange pictured below?  Would this be a "make Up" compressor that is only used in higher demand time periods?

I have had the pleasure of seeing this motor through two rebars in 5 years.  When the rotor finally failed, the tell tale audible beat could be heard outside of the boiler house.  Short story is this motor is a make up air compressor and second shift Ops always elected to shutdown and start up the unit when needed, not just unload the compressor.  If my memory serves me correctly 14 starts in a day was the max, usually avg 7-8 a day, across the line of course.  Do you know the number of starts or duty cycle of the unit, just some food for thought regarding your rotor suspicions. 

As a motor shop guy, I was never out and collected data under load at this facility when the rotor had failed, only unloaded shop data.

I have not used the parameters that Jim Crowe suggests in this post, review and see if it is something your could add to your route on the next visit to help maybe?

https://www.machineryanalysis.org/post/show_single_post?pid=1291866283&postcount=11&forum=473361





photo.jpg 

Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,244
Reply with quote  #11 
Rusty,

There is often a change in load which I was told by a compressor guy with years of VA experience can be due to a change in humidity. More humidity, higher density.

It usually shows up as an increase at VPF of the primary stage (around 20 x running speed) but I never looked to see if it effected the 2 x lf on the motor. I don't have any records on humidity to confirm it but we do those compressors twice a year and as I recall, it's almost uniformly higher in the summer.
KVib

Member
Registered:
Posts: 7
Reply with quote  #12 
Set your 2140 in live mode. Set your fmax to 10khz and see if there's any high frequency peaks.
__________________
Regards
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,607
Reply with quote  #13 
KVib, I do the lubrication on these motors so I have a special point with an Fmax of 10 kHz that is collected every month. If I see (or hear) high frequency content, I lubricate the bearing, and then collect the point again to confirm/document. I don’t recall that point showing anything, but I’ll look again. And you’re right about “live mode” - that’s a useful tool which I don’t use often enough. Often something that’s intermittent just doesn’t show up well in a “snapshot” spectrum.

But the noise I’m feeling is not high frequency. It feels more like the slightly raised noise floor seen in the multi-spectral plot, from 200-600Hz.

__________________
"The trend is your friend"
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.