Skip to main content

Dear peer,

We have boiler feed water pump driven by HT motor of rating 500 KW, 32 FLA, SKF 6319 at both end with locating bearing at NDE and Non locating at DE. Pump is coupled by gear coupling.

As per previous maintenance history we get high vibration 4.0 mm/s (RMS) at Motor DE axial at 1×BPFI (~14500 CPM) durung motor no load trial just after overhauling (1X is very low, near about 0.7 mm/s). All other points vibration are well bellow 2 mm/s (RMS). we did not get BPFI frequency in radial direction. This problem persist since last long and even after bearing replacement for several times. In coupled condition, we observed same scenario. Sometime BPFI peak goes down upto 2 mm/s (RMS) but it never vanished.

This time after overhauling we get 4.5 mm/s (~14500 CPM - 1X BPFI - SKF 6319) at motor DE axial. No BPFI peak in radial direction. We did some activities and we have two observation as below.

1. When we loose motor drive side mounting bolts axial vibration reduced to 2.3 mm/s (Dominated 1x BPFI). After loosening of all four bolts we get vibration 1.9 mm/s (Dominated 1x BPFI). When we retightened bolts vibration increased to 4.5 mm/s.

2. We carried out bump test on motor DE end bell. We got a frequency at 14466 CPM.

Please explain above behaviour of motor and suggest solution.

Thanks and Regards

Ikram Patel

Tags: Natural frequency, Bump test, SKF 6319, Motor No Load, Solo run, Hold down bolts, Axial vibration, Mounting bolts, BPFI

Replies sorted oldest to newest

SKF 6319 is a deep groove ball bearing, so the inner race/ring defect must be small and have a specific location that is affected by rotor axial position and radial/axial load.

6319 - Deep groove ball bearings | SKF

Was motor overhauled with new bearings or just work on pump?

Have you considered the possibility that cause of premature bearing defects are related to installation procedure, lubrication, or there is stray motor shaft current that causes electrical discharge machining (EDM)?

A gear coupling can transmit axial load to bearings if the lubrication is poor, if teeth are worn, shaft end gap is incorrect, and if pump and motor shafts are misaligned.

The resonant motor DE end bell is just bad luck that amplifies the forcing frequency from the bearing fault. You could alter the resonant amplification by adding mass, stiffness or damping. This would reduce vibration level, but it would reduce the early detection of a bearing fault.

Do not modify the end bell structure until you identify and correct the root cause of the premature bearing fault/failure.


Last edited by Registered Member

Thank you @Registered Member for quick responce.

I would like to add that motor was overhauled with new bearings.

As of now 4.5 mm/s vibration is at motor no load condition (without coupled) trial. Hence coupling effect can be eliminated. Near BPFI frequency appear from very first minute of motor starting. We have replaced bearing multiple times but same problem persist. Removed bearing does not show any defect.

We have experience that due to angular soft foot 1X vibration varies as motor base bolts loosening/retightening work done. In this case during motor basebolts loosening/retightening work near BPFI frequency varies. I would like to know that Is it possible that stress induced at motor leg due to angular soft foot can influence near BPFI frequency/ Natural frequency at end bell without affecting 1X ?

Please suggest any temporary method to modify stiffness, damping, mass of end bell for testing.


Okay the dynamic force is not from a drive end bearing inner race/ring fault. Assuming you have 50 Hz power frequency and that it is a two-pole motor, the vibration frequency is close to 5X shaft speed. The excitation force would not be electrical, so mechanical looseness is expected. The looseness could be within the fit of the bearing or another part of the motor structure. The fact that the vibration level changes from running with motor uncoupled to running motor coupled and with load suggests that the natural frequency has shifted, or vibration amplitude has reduced from less looseness or structural damping.

Some motors have a wave washer on the DE bearing to provide some pre-load.

I cannot make any suggestions for detuning the natural frequency by modifying the DE bell housing without seeing one or more photos and a scale drawing or sketch.


Last edited by Registered Member
@Registered Member posted:

@Registered Member please share your thought.

Since you can loosen a motor foot and get almost a 1/2 reduction in vibration and loosen all four feet and get in excess of 1/2 reduction, I would be concentrating my efforts on the foundation.  Keep in mind that a gear type coupling when misaligned generates axial force.  In particular, try to understand what loosening (and tightening) the hold down bolts does with respect to misalignment.  For instance, have you put dial indicators on the corners when you loosen the bolts in an attempt to quantify the possible distortion of the motor casing?

Last edited by Registered Member

Thank you @Registered Member for your valuable comment. 

Yesterday we have worked on foundation and found that there is problem in skid level. We have done temporarty adjustment by putting shims. Overall vibration at motor drive end horizontal is below 2 mm/s with all four bolt tight. So this one might be reason for high vibration.

I am still wondering why this does not excite 1X vibration ?

Thanks and regards

Ikram Patel


"I am still wondering why this does not excite 1X vibration ?"

Perhaps just a misunderstanding of terminology. Vibration at 1xSS is the result of an excitation force such as rotor unbalance or shaft misalignment. The foundation does not excite vibration, but it can amplify the vibration response. Several factors with the "loose motor base" can occur, such as change in natural frequency, change in damping, change in vibration transfer from motor frame into foundation, change in hammer contact (secondary vibration force) between two loose surfaces. If your effort has reduced vibration level to your satisfaction, then you are done. If you want to under the root cause of the original vibration levels, then more measurements and test would be needed.



thank you @Registered Member, @Registered Member  for valuable comment and sorry for late reply.

in continuation with above problem we have taken motor solo run trial (Without pump coupled) on 25.04.2022 and then coupled with pump. Overall vibration again increased drastically on motor both end in axial direction at 1X and 2X. interestingly without any work done and without pump stoppage this vibration again reduced with dominant peak at around 12K CPM at motor NDE axial and at 14560 CPM at motor DE axial. pump side overall vibration in increasing trend. Please note that pump has DE side journal bearing and NDE side Journal + double row angular contact bearing.

This time motor side vibration is totally unpredictable. Spectrum dominant frequency, Overall vibration change without pump stoppage and maintenance activities.

Please refer attachment for better insight.

Attachment 1 - Overall vibration from 25.04.2022 to 29.04.2022

Attachment 2 - Multiple spectra of motor bearing indication change in dominant peak and vibration from 25.04.2022 to 29.04.2022

Attachment 3 - Motor + Pump bearing spectra taken on 29.04.2022



To make it easier for the readers, it's better if the spectrum x-axis is expressed in order while the speed is accurately configured.

WRT the intermittent/unstable vibration, one of the possible reasons is the unstable lubrication (oil flow, dirt in the lubricant, unstable grease location, etc.). Another possible reason is the thermal effects either from the process side or from the metal expansion/contraction. Clearances can be sensitive to temperature due to the dimensions of the machine's components; parts can get loose or too tight.

Do you have a way to plot the vibration and temperature in a trending format (e.g. every minute)?

If I were you, while the machine at the site, I would carefully inspect the old bearings to identify any defects, even at the early stages. At the shop, I'd carefully examine the areas at both housing OD and shaft ID sides.

Regards- Ali M. Al-Shurafa


Add Reply

Link copied to your clipboard.