vibration

[32c4]

I need documentation or help for a problem of vibration.first a will explain to you that I have this problem in a vertical pump driven by a motor at speed of 3000rpm and power 75KW ,motor with 2 bearings upper and lower and pump bearing with 2 shaft assembled together.
the problem happens at 24.5 Hz we have a peak of 9mm/s (it means a 1/2 fundamental frequency).
the high point of vibration is at the upper bearing of the motor.And vibration at lower bearing of the motor is minus than the upper,
the temperature of the lower bearing is 85°C.
vibration level at the bearing housing of pump is less,1.7mm/s
but always 1/2 fundamental frequency which appears.
two shaft assembled are 2.5m
1m for hydraulic and 1.5m for transmission.
no-load motor test vibration all signiture of vibration disppear .


mahiddini mohamed
thanks

 

[electricpete]

As you probably know, vibration at 1/2 times running speed often signals a looseness of some kind.

Check for looseness of things that area easy to check. Motor holddown bolts, things attached to the motor (like terminal box). Especially if the 1/2x shows in one direction and not the other. It's probably not likely you'll find these things (since vib went away uncoupled), but still worth checking first imo since it's easier than the other things.

Some motor thrust bearings (like single spherical roller bearing or single/tandem angle contact bearing) rely on a minimum downthrust to keep the bearing engaged and lose their radial stiffness if the pump upthrusts. I have seen this give non-syncrhonous frequencies not exactly 1/2x, but I wouldn't be surprised to see 1/2x. Most (not all) vertical turbine pumps can give upthrust only when they operate at very high flow rates beyond BEP toward runout. If you have these bearing types and the 1/2x vib pattern reduces as flow decreases you might be more inclined to suspect this.

Vertical motor driven pumps often show vibration at somewhere less than 1/2x when the pump bearings degrade (we have one that shows at around 0.35x running speed). I can't say I've seen worn pump bearings show at 1/2x but in theory I can't rule it out.

There can be a variety of components on the shaft, pump rotating components, pump stationary components, couplings that can be loose. Those are a lot harder to check without inspection.


75C on bottom bearing at first glance sounds high. … measured on bearing housing or bearing? Greased bearing? Does this motor have thrust bearing on bottom by any chance (a few do). If radial bearing at 75C it suggests a problem of some kind which may or may not be related to your vibration.


=====================================
(2B)+(2B)' ?

 

[JJPellin]

You did not state what type of vertical pump you are dealing with. I would assume a vertical turbine pump with no true bearings in the pump (only bushings) and a rigid coupling. Please correct me if I am wrong.

Vibration at one half-speed can often indicate a shaft whirl/whip condition. We have seen this in vertical pumps in liquid sulfur service. When we had a detailed analysis performed, we determined that the shaft bushings were inherently unstable and the shaft was going into a whip mode at about ½ running speed. We changed the bushing design to one with three tapered pressure dams that basically form a triangular bearing that is inherently stable, even for a vertical shaft orientation. I could provide references for the particular bearing/bushing design that we used if you are interested.

Some would argue that a smooth bore bushing around a vertical shaft is always going to be inherently unstable. The instability may not be demonstrated in the vibration data only because an imbalance or non-concentricity side loads the bushings and makes them more stable. If this is the case, the more perfectly aligned and balanced the pump, the more likely it is to be unstable.

It will be easier for us to provide good suggestions if you could give us more details about the pump, motor and process.


Johnny Pellin

 

[Ripz]

Is your vibration exactly 0.5X? Usually you hear of vertical pumps getting shaft whip conditions becasue of increased bushing clearances but this usually manifests as (approximately and varying) 0.47X

If it's looseness (extreme cases) you'll get exactly 0.5X but with multiple harmonics of this RPM.

 

[32c4]

First, I tell you that the motor alone without pump doesn’t vibrate, but coupled will vibrate,
In none drive side or end of the motor only, at 1/2 rotation frequency (25Hz) we have a peak of 9mm/s, on the spectrum. The other side of the motor (DRIVE END) nothing and also on the pump bearing nothing or value of vibration is good.
BEARING OF THE MOTOR AT THE NONE DRIVE END: NU314C3.
BEARING OF THE MOTOR AT THE DRIVE END: 6314C3.
Pump 2 ball bearings X disposition:7313BECBJ
Motor power: 75KW.
PUMP with 2shafts assembled together one hydraulic and the other transmission.
pump shaft is long of 1m and half for the transmission
Hydraulic is 1m long
we are using METASTREAM Type TSKS COUPLING WITH
Disc pack or membrane pack.
thank you

 

[Tmoose]

Hi mahiddini

Have you done a "bump test" horizontally in 2 perpendicular directions at the top of the motor?

http://www.mobiusinstitute.com/tips.aspx?id=679

regards,

Dan T

 

[32c4]

Hi Johnny
I believe it is a very complex subject above my head and you have suggested that you could provides references for particular bushing design .
yes,Iam interesting can you explain me more
Thank you for sharing your knowledge .

 

[Ripz]

Your pump bushings may be worn. Its not your motor. Youre only getting the pump vibration being transferred to the motor.

 

[electricpete]

To 32c4 - is the frequency you're seeing exactly one half running speed.

To everyone else - is exactly one half running speed a common frequency for worn bushing-style bearing to exhibit?...

As I mentioned above, we have seen worn bearings on vertical pumps show up at less than 1/2 running speed but not exactly one half running speed. But I've been wrong before and open to learning.


=====================================
(2B)+(2B)' ?

 

[JJPellin]

The design change I was referring to was actually presented as a Case Study at the 26th International Pump Users Symposium. The case study materials can be found as follows:

http://turbolab.tamu.edu/proc/pumpproc/P26/Case Study 08.pdf

In that case study, they found that they were exciting a shaft critical speed and a structural resonance at 1/2 running speed. The excitation was coming from the unstable plain circular bushings. This matches our experience very well. The first author of this paper was hired to assist us with our problem. We ended up implementing the same solution with tri-lube bushings.



Johnny Pellin

 

[32c4]

Hi friend's
it's exactly one half running speed.
added information
pumped liquid: propane
discharge head :114m
quantity:292.8m3/h
speed:2950rpm
pump type:16WUC-3+ind.flowserve.
If I REDUCE THE SHAFT DEFLECTION , BECAUSE THE RUN OUT IS ABOUT 8/100 OF ONE METER AND HALF AND THE RECOMMANDED IS 3/100 BY METER

friendly mohamed

 

[electricpete]

Do you mean you have 8/1000 meter shaft runout and your limit is 3/1000?
Even the 8/1000 m ~ 0.3 inch TIR would still be quite high.
The runout is measured with dial indicator or prox probes?

Rigid coupling?

Seems you have found a clue….

I'd think first steps of investigation might be uncouple, inspect the coupling, and see how the pump drops. i.e. does it drop axially the proper amount, and how centered does it end up compared to motor shaft (dial indicator base on motor shaft indicating off pump shaft).

Maybe see if it can aligned and then recoupled and see if still has that huge runout goes away.

That's mostly the easy stuff. If the runout is still there then I'd be inclined to think the pump should be pulled for inspection / repair.

I don't work on pumps that much. Others may have better suggestions here.


=====================================
(2B)+(2B)' ?

 

[Tmoose]

is this:
- a new pump that has attracted attention during acceptance testing with this vibration
- a freshly rebuilt pump that has attracted attention during acceptance testing
- a pump that has been in service a while and just developed this vibration
- something else altogether?

and, have you done a "bump test" in 2 perpendicular directions at the top of the pump in the plane of the vibration that is concerning?

is the "runout" deflection being measured close to the coupling, on the motor shaft and on the pump shaft?

Like EPete said, numbers like 0.004" or 0.1 mm ( m/10,000 ) what I'd expect for a runout tolerance.
There are several things that can cause excessive runout, and excessive runout can cause vibration (although not immpossible, not really likely at the ~ 1/2 running speed you have).

 

[32c4]

Hi
excuse me
from the instruction manual of the pumpwe have
The set point for the maximum deviation is of 0.03mm per meter length of the shaft.
And I have 0.08mm for one meter and half the length of shaftpump.
after disassembling the pump ,we noted that also bushings are out off tolerance and wearing rings are damaged.

In a friendly way
for you

 

[Tmoose]

Does the METASTREAM Type TSKS COUPLING use a single pack or shims, like this?

http://image.made-in-china.com/2f0j...ling-Flexible-Coupling-Diaphragm-Coupling.jpg

Or,2 packs separated by a spool piece? Like this -

http://image.made-in-china.com/2f0j00CMsapAYFrcbD/Diaphragm-Coupling-J6-ZJ6-.jpg

======================
Without seeing more of the manual I suspect 0.03mm per meter ( 0.0011 " / ~ 40 inch) is a straightness spec for the un-installed shafting.
Measured something like this.

http://www.roton.com/images/33.gif

===============


Problems created by The installation can cause shaft runout even when starting with "straight shafts". If shaft ends butt inside rigid couplings and the machining is poor, or a piece of grit is left between the faces the shaft assembly will be "kinked" when the coupling is tightened.

Diaphragm coupling bolts must be tightened with the shafts in good alignment and the disks "neutral.". Otherwise they will be trapped in a bent condition and behave like fairly weak but bent shaft, which they are.
Badly machined hubs can do something similar.

 

[32c4]

Hi
our's help has been greatly appreciated.I will tell you if there's news after assembly the pump and test .
see you
greetings

 

[Smercer]

Looking forward to the results 32c4
PdM Specialists, LLC