Hi All,

Do you mind to share on different technique on how to analyze sleeve bearing at early stage of fault if only casing measurement is available? OR is it really possible?

We are not able to effectively use casing measurement data (or at least lacking confidence) to detect sleeve bearing problem at its early stage.

Some techniques We normally use:
- waveform pattern to detect abnormal G-s spike which we think not normal (we monitor the trend),
- increase of multiple harmonics of TS
- look for raising noise floor
- using live view to detect possible oil whirl (if we suspect some problem).

Is there any other method which we can add to the list?



Best Regards
Arief

Dear Arief

Two very good methods :

1- Monitor oil temperature.That is the most significant clue that your sleeve bearing is going to fail soon.
2- Try to install a proximity probe against part of the shaft near the bearing.

Marko leo

quote:

eary stage of fault

This is a term applied to rolling element bearings. It's applicability to fluid film bearings is questionable.

Hi Arief,
I had some confirm result with PeakVue technology from CSI. I configured casing measurement on sleeve brg with speed from 163 rpm to 900 rpm with the Frequency calculator under "PeakVue assistant" folder. I notice an increase in waveform Peak to Peak on the trend even if the spectrum didn't show any peaks. Waveform analysis shows 1 x TS impact. After removing the brg we confirm that the residual gap was at least twice the original and Peak-to-Peak amplitude felt down after brg replacement. Temperature trending doesn't give us enough warning and the brg often failed before replacement maybe due to low speed and ambient temperature influence. Prox. are more efficient in monitoring sleeve brg but I feel that Peakvue technology and temperature trend are cheap ways to achieve reasonable result with sleeve brg monitoring depending on the machine.

Best regard,

See:

http://vibrotek.com/article.php?article=articles/brnvi97/index.htm

and

http://vibrotek.com/article.php?article=articles/brn97/index.htm

and

http://vibrotek.com/article.php?article=articles/intelect-eng/index.htm

As Bill mentioned, sleeve bearings have a tendency to fail rapidly. However, the rate at which they fail depends on speed, load and distribution of it, lubrication and vibration.

How does your machine fit the above descriptions. If your machine is a high speed machine greater than 3600 rpm then you don't have a real chance of predictive analysis. On the other hand, if it's a 1800 rpm machine with large sleeve bearings like Dodge pillow blocks you have a better chance to see "failure stages". All this is possible providing the sleeve bearing is not stressed by preload, bearing misalignment, lubrication etc.

And of course lub monitoring is another important piece of the puzzle that may improve your monitoring capability for sleeve bearings.

Monitoring early stage! Early stage or start-up isn't a continual monitoring process but more of a vibration acceptance test utilizing vibration analyses rather than PdM route.

Hopefully you have valid spec info such as bearing clearance (tilting pad???) for a standard sleeve.

Mass is a factor concerning amplitude analysis for your cap data. Cap temp and oil temp should be monitored as well. I've caught new start-up and called for shut-down and inspection at 0.03 IPS for a rub and cap temp was going up as well.

All, thanks for responses.
Behind our question, actually are non-critical machines suchs as pump and turbine with sleeve bearings and running speed between 1000 to 3000 CPM. Due to cost and quantity of these units, we can not justify to have protection or online system, therefore heavily depens on our routine monthly casing vibration measurement.

From the casing vibration readings, Other than overall value, We are trying to find out if there is 1 or 2 specific parameters which is important enough to be trended and get our attention for further analysis. And how to set our analyzer to capture it.

We are currently using CSI 2130, and will be interested to know if anyone of you have special setting for sleeve bearing application.

So far, parameters your response :
Marcel B: Peak (or Peak to Peak) Waveform, 1XTS period
Duncan Carter: Demodulated Spectrum 1xTS and/or its harmonics
Other: Oil temp/analysis, cap temp

We suspect the demodulated, envelope or PeakVue (FFT and Time domain) type of measurement could be helpfull to indicate the sleeve bearing. Unfortunately, by default this parameters are not assigned if we specify the bearing as sleeve type. Another factor, due to random, intermitent or short duration event between sleeve, lube and shaft, we could easily miss the events due to selection of averaging and waveform time-span.

Or we really have to manually see the data for to each point when analyzing sleeve bearing?

Arief2072,

"We suspect the demodulated, envelope or PeakVue (FFT and Time domain) type of measurement could be helpfull to indicate the sleeve bearing. Unfortunately, by default this parameters are not assigned if we specify the bearing as sleeve type."

I do not use CSI, but Rather IRD. However, all through working for the company (10 years) we were told that Spike Energy would not do any good with sleeve bearings, since they had no high frequency content. It appears from your statement that CSI has decided the same thing, but doesn't allow you to see for your self when picking some automated setup.
If I hadn't played around with spike energy and sleeve bearings, I may have believed what I was taught and never found out that it is a VERY GOOD sleeve bearing measurement. You have to have your filters set up properly, but you can certainly see the increase in harmonics as the bearing wears. A normal one usually has three or four harmonics, a worn one may have 8 or 10, a wiped one has high noise floor and harmonics out as far as your F-max.

If I was you, I would play around with it (peakvue). Use the rule of setting your filter down about where your F-max is. On a 1200 RPM big ID fan I don't go out past 30k, on an 1800 RPM pump, I use two readings, one looking for vane pass, etc. the other down around 30k with gSe.

Good luck,

Dave

Here is a sleeve bearing (@870 rpm) that recently enlarged it's clearances. The slower speed gives us a nice time line to look at. All vibrations were really low. Temperature was the main indicator of a problem. No work was performed. Coupling spacer needs adjustment.

Pre wipe info
12/20/2007
1X and 6X spike <.05 IPS
6X = new coupling
Temp @160F

Post wipe info
12/24/2007
1X <.05 IPS
6X gone
Temp @140F

We should fix the spacer next outage. It's almost a year from now. My challenge will be to ensure it stays in the plan. I think the higher temperatures during the summer run will keep it on the radar.

Happy new year!

Is this bearing metal temperature?

What is the inlet oil temperature? This helps with the trend.

Is 140 a believable temperature? To what do you attribute the low temperature?

Hi
I just want to add some more to what other members have discussed. Usually the sleeve bearings last long if designed properly and proper lubrication is maintained while in operation.The wiping type of failure results in increased clearance of bearings(purely sleeve bearing not tilting pad) which reflects in 3X vibration. Keep track of 3X and 4X harmonics if you start suspecting bearing deterioration.Enveloping technique can be sometimes very helpful but the bearing would have already crossed the 3rd/4th stage of failure when u can detect pulses out of impacting.But Oil Monitoring in terms of Wear Debri Analysis is positive technique for Journal bearings
bye

Is this bearing metal temperature?
Yes it is.
What is the inlet oil temperature?
N/A This bearing has a small oil bath. @1/2 gallon.
Is 140 a believable temperature?
Yes. We have 8 same design/process motors all inboard bearings are 120 - 150F.

To what do you attribute the low temperature?
The motor cooling fan pulls in across this bearing (as well as the outboard) and discharges out of vents on the side.

Why did it drop to 140F?

When the bearing wiped, the babbit moved around the bearing. The softer metal was removed completly or deposited away from the tighter areas. The oil is now allowed to enter the larger clearances and shaft is enjoying the chance to ride on a bigger oil wedge.

PS:No vibration energy at 3X or 4X.

This message has been edited. Last edited by: CBM Novice,

quote:

The oil is now allowed to enter the larger clearances and shaft is enjoying the chance to ride on a bigger oil wedge.

So, why didn't the bearing designer simply increase the clearance? I don't think this is the way the physics works.

I would have concerns about collateral damage, like journal damage, in the future.

Can you get a temperature reading on the oil sumps? This might serve as an indicator related to the oil temperature in the sumps.

It doesn't sound like it could have been a complete bearing wipe. One thing that can happen is that the bearing profile starts to resemble the journal when you start wiping. This leads to the inability to support the shaft load but temperatures would not decrease.

Hi Arief,

If you need protection system or online protection for sleeve bearing, you better use temperature sensor. Vibration measurement in casing will be too late to protect your system.

Sleeve bearing fail rapidly, when there is lack of lubrication or no cooling.

We had been in this situation before, when we change the damage sleeve bearing with new one, as we suspect lack of lubrication.

But the new sleeve bearing only last for 10 minutes and damage again during running test. The vibration measurement showed good until some smoke came out from the sleeve bearing.

The problem was the cooling water instrument's valve was jammed. The instrument showed open in control room, but actually no water was flowing in the field.

Regards,
Bimo

quote:

All, thanks for responses.
Behind our question, actually are non-critical machines suchs as pump and turbine with sleeve bearings and running speed between 1000 to 3000 CPM. Due to cost and quantity of these units, we can not justify to have protection or online system, therefore heavily depens on our routine monthly casing vibration measurement.

CBMNovice - thanks for sharing the case study. It sounds like an interesting one. Is the spacer coupling believe to have been involved in the problem somehow? Maybe the motor rotor was axially mispositioned so that there was contact at the shaft shoulders outside the bearings (intended only for intermittent contact during uncoupled run)?

Happy new year Ya'll!

quote:

So, why didn't the bearing designer simply increase the clearance?

I think the original design was fine. The bearing was OK until the new coupling was installed.

What is a complete bearing wipe?

Pete - That's the thought I'm having. OPS describe the motor as moving 1/2" when it's secured.

1. The electrical center scribes are missing on the motor shaft.

2. Without the scratches I'm assuming the check was not performed.

3. Without the electrical center defined, the correct spacer size for the new coupling can not be known.

4. 1/2" is all the axial movement available on the motor. The 3 scribes can be seen on another motor (same type) that is secured.

This message has been edited. Last edited by: CBM Novice,

Happy New Year!

Still an interesting case. Intermittent hot running could be due to improper setting of coupling gap. But if the change in position during shutdown (and I assume the machine remains coupled), if that is real, then it suggests something else. Normally the axial position of a horizontal sleeve bearing motor is controlled by thrust bearing in the driven equipment (through the coupling) which might have clearance somewhere in the range of 0.005 - 0.030". Do you have any more details on the coupling and driven equipment thrust bearing? It might be worthwhile to investigate the condition of that driven-machine thrust bearing.

Hello gentlemen,
I worked a lot with sleeve brg synchronous motor that had separate enclosure (brg and stator were on separate base). We came up with an installation procedure since no marking were available. First you must install sleeve and shape the lower part of the Babbitt brg to fit the shaft with Prussian blue and hand held brg knife. This will ensure that there is enough surface contact to support the oil film and the machine load. It will prevent start-up failure. The motor has to be tested running without the coupling so you can measure the position of the shaft axially (magnetic center). Stop and lock motor and check mechanical axial play within the brg. Compare both and make sure there is sufficient space for the stator to move at its magnetic center. Ideally mechanical and magnetic center should be exactly the same but we usually accept ± 1/16". From the magnetic center measure the residual gap of the coupling and cut or install shim accordingly do not forced the stator out of its magnetic center. It's better to move the driven machine to adjust the coupling gap. Forcing the stator out of the magnetic center always end with problems. We had numerous failure of our motor due to improper installation and once we got it under control everything ran very smoothly afterward.

Best regard,