Hi,

Here are the specs of the motor on which site balancing was attempted 4 days back.

Machine specs:
2-Pole motor
3000 rpm unit
Rating: 450 kW

I am hereby attaching the study report showing HVA spectrums, balancing readings, coast down data, bump tests results, history details and unit sketch.

client requested to take up site balancing of this motor.

motor was suitably clamped.

foundation intact. high vib levels on Motor NDE vertical (26 mm/sec RMS).

spectrum shows pure 1X component, time data does not look like its a balancing problem.

an attempt was made to balance the motor.

mass correcion was provided near NDE on fan.

Motor did not responded for mass addition

NFT indicated 51.9 Hz... close to running frequency

coast down data shows sudden drop in vib levels

Please see the report attached

can someone help. thanks in advance.

Madhu

vibration_measurement_report.pdf (194 Kb, 89 downloads)

Hello
This does not look like a balance issue check the rotor for eccentricicty for one as this will cause your levels to drop once the power is cut. You waveform also does not confirm a balance problem alone You may also be seeing a resonance issue. This is just my opinion and is subject to discussion and correction as others here i know have a more indepth knowledge of balacing machinery.Check these items and try to eliminate them first then move to the balance if there are no other influences.

Madhu,

So if I understand this correctly the motor was running ok for a few days then started to vibrate. That is usually not a balance issue unless you lost some mass. Any indication of this?

Your coastdown starts low at 3000 CPM then increases and then falls off. Is this what happened?

Last question is: What was the angle location of where you installed the trial weights?

Madhu
Do you have a time waveform of this motor immediately after shutting power?
I agree with you re: balance. I learned a long time ago that if machine does not respond to balance shots - you dont got a balance problem.

Its possible you may have some "looseness" internally. Was this motor recently repaired?
There are electrical checks you can make. I use an All Test device to check integrity of motor.

Also, I may be wrong here but does it seem that 75 RPM slip is kinda high for an uncoupled motor?

Regards
Jim P

Madhu,

You did not provide impact test results for the NDE but coastdown data suggests that you may have a resonance right there. That is why balancing was not effective although it does sometimes. Move the resonance away and try balancing again.

David

Hi,

The motor was running at 2925 rpm when the data was taken

trial weights were added at exactly near the reflecting tape (0 deg)

but its of no use the motor is not responding for mass correction

the coast down data curve starts at 2925 rpm and not at 3000 rpm

there are no harmonics in the spectrum... phase was consistent....does it still be suspected for looseness

no repair work was carried out on motor, yes bearings were replaced... but will it shift the natural frequency...???

time waveform was observed immediately after cutting off the power... it went down gradually and not immediately...!!!

impact test on the NDE side with sensor in horizontal & vertical direction did not show any freq's which are close to running freq.

but i am getting the 51.9 Hz peak in bump test on both Motor DE and NDE with sensor mounted on the shaft in axial direction.

cannot shift resonance... the motor is sitting tightly on the foundation with all bolts intact.

does this indicate any crack in the rotor bar/slots/assembly where the resonance is appearing all of a sudden...????

Madhu

quote:

Based on the history data collected at site, the motor was in operation for
about 2 to 3 days, in coupled condition, with blower assembly without any high vibrations.

No vibration for the first 2 to 3 days and then all of a sudden the amplitude jumped up? Is that what happened?

quote:

There was no predominant change in the vibration levels of both motor DE
and NDE even after several attempts of mass correction (refer table-4)
• A total of around 145 gm was added on the NDE fan assembly for which the motor did not respond
• Normally a high speed unit of this capacity should respond for a very minimum amount of mass correction (around 20 to 30 gm) because of the higher centrifugal effect

This SOUNDS like the problem IS NOT imbalance, but:::::.

The "power off" plot shows 1 of 2 things to me,
The motor is running directly on top of its critical frequency or just a few rpms lower (which should have shown from the start up rather than 2 or 3 days later) or there are rotor problems (broken/loose bars) which could have happened after the start up, especially if the motor had been restarted several times after the first startup or even while running constantly the 2 to 3 days.

If it is running on its critical, the amount of weight to make a balance change would be much less than if it is not running at its critical. So 145 grams grams would have been WAY TOO MUCH weight IMO (depending on the radius used) even 50 might have been too much.

I see the phase only changed about 60 degrees total on all balance attempts, but the amplitude stayed in the same area. Could this be a coincidence or normal for what the real problem is?

I ran your numbers through a balance program, using 50 grams, position at "0" and CW as direction and came up with 138 grams at 231 degrees. Is this where the 145 grams was placed? Running at or near a critical, the angle and weight needs to be as close as possible to what is calculated, IMO.

Can you zoom in on the 1x with high resolution and see any sidebands of slip X # poles. I agree that a 75 rpm slip unloaded is strange. Sounds like something is slowing the motor down (loss of power) if the fan is not connected to it.

OMOAICBTW

Did you check the cocked bearing condition, Bearing designation before and after(Some times instead of C3 people will go for C2 or Normal clearance bearing)
Could you check this and update me??
Gurusamy

Hi,

I am getting several queries from everyone and hope i had answered all the points with suitable data on this particular problem. but i am still in dark regarding what may be the exact problem which is causing the motor to vibrate at 26 mm/sec and not responding for balancing. is it resonance or rotor/stator bar cracks or something else...??? what should i tell to customer... where should i tell him to put his hands on the motor for repair/rectification of the above problem.

Ralph i had attached the zoomed view of the data you had asked for. sorry the resolution is only 800 lines. i do not have the raw data, yes the correction was asked by my analyser at 231 deg and i have added the exact mass in first shot on the cooling fan near motor NDE.

here is the motor name plate details:

3 PHASE SQUIRREL CAGE INDUCTION MOTOR
KW :400 KW
FRAME : LD 450 L
RPM : 2979
AMPS : 84 A
VOLTS : 3.3 KV, 50 HZ,3 PHASE
ENCL : IP 55, S1 DUTY
CONN : STAR
BEARING DE : 6219 C3
BEARING NDE : 219 C3

Hope it helps,

Madhu

vertical_sign.ppt (96 Kb, 35 downloads)

Madhu,
Look again at the bump test at NDE (page 6). There is a little hump around 51 HZ. If you blow the plot up vertically or use velocity units you will see it better and even will be able to calculate AF.

The coastdown data, as mentioned above, also points to a resonance which is slightly above the running speed. If phase data is also available then a conclusive decision could be made.

David

Still not adding up to me. Thinking out load here.

Ran good for a few days then bad - does not sound like balance,

Some data looks like you are running as resonance but the balance attempt should have been very sensitive to the weight and it was not.

I see you added weight on 1 end only. Is it possible that the weight location is a node point and therefore has very little effect? Balancing a big motor on one end may not be very effective.

Ron
I agree with you.
At first resonance seemed logical EXCEPT it doesnt explain why it ran OK for several days prior.
That statement is key. Something changed in the system response.
Same argument for balancing - If it was a balance problem, it never should have run ok for a few days....or at least if something broke off, it then should have responded to balance attempts.
Still thinking out loud with you :-)

JP

Logic should be driven by the data and not by assumptions which could not be tested. At this time data supports possible resonance. It could and should be proven or disproven by testing. After that one can move ahead.

The fact by itself that motor ran well for a few days can't support neither presence of resonance, nor absence of resonance, nor unbalance, nor anything. Things happen randomly. It is entirely possible that motor was running close to resonance from the very beginning with minor unbalance until something broke off and now resonance became evident. But this approach shouldn't be driving troubleshooting efforts IMO.

Although this is not the main problem here, I agree, balancing a motor at one end is not proper.

Hi,

Thanks for all your inputs. i do agree that we need to carry 2-plane balancing here. but unfortunately there was no provision near the motor DE. the only option avialable was near motor NDE and that is on the cooling fan and this is the place where we have high vib levels. that's why this whole exercise.

everybody's statement boils down to "THE MOTOR WAS GOOD EARLIER RAN FOR A COUPLE OF DAYS AND THEN THE PROBLEM STARTED"

yes that's correct and the vib levels increased thereafter. my feeling is if the resonance factor shows up all of a sudden then there must be some change/crack/damage in the rotor assembly which suddenly started exciting the resonance which was sleeping earlier. just a feel. otherwise what else could be the main reason. i am not finding any other reason beyond this.

David I have attached the bump test results of motor NDE in radial direction in velocity mode. sorry i do not have the phase data.

Madhu

bump_test_near_NDE.ppt (74 Kb, 31 downloads)

Madhu,

I think it is a resonance. I have estimated the amplification factor and came up with AF=11. this is a good number for a resonance to do nasty things. You can calculate it more accurately from the plot. I'd carefully check the DE for the same matter and then start making assumptions ( possible causes) followed by testing.

Madhu,

Why is this motor running with such a high slip speed (75 rpm) if it is not connected to the fan? Or is it connected to the fan? I seem to not be able to find in any of the replies where it specifically says the motor is running uncoupled. I do see on the attached report that it is running "UNLOADED" but does this mean uncoupled from the driven part?

If it is uncoupled, then the same question is asked, why is it running so slow (2925 instead of ~ 3000 rpm)? If it is uncoupled then resonace is not the problem. Resonance would not cause such a drastic loss of speed.

SO IS THE MOTOR RUNNING UNCOUPLED OR COUPLED? Or is it a variable speed motor and the 2925 rpm speed is the speed it would run if it were coupled to the fan and this is why there is such a high slip frequency, because this is the speed at which someone has chosen to balance the motor? The "tag" says it should not run lower than 2979 rpm. So again, why is it running so slow?

You said there is no access to a balance position on the DE of the motor. Can you not use the coupling as the balance plane on the DE? But then again, it does not look like it is a balance problem based on the hugh slip speed and the rapid drop in amplitude when the power was cut off. True, if it is running at resonance, a small amount of imbalance will cause a hugh amplitude to be attained. But I am concerned with the 2925 rpm speed on an unloaded 3000 rpm motor.

Some have suggested maybe something has "broken off" after startup.

Have you got any data when the motor was first installed 3 day prior to the problem?

One more question. Is this motor outside or inside a building and is the air cold or hot and is the fan cover on or off after each weight adjustment was made?

Only my opinion and I could be right or wrong, 50% chance either way.

Just my two cents.
Witnessed a motor do the same thing with 1x. Took the motor apart and put the rotor in a balance machine.
It was then that everyone noticed a bow in the rotor shaft. Dial indicator readings confirmed a perfect 'U' bow of over 0.018".
Ran beautiful for two weeks. What happened?
Took a good nights sleep for the smoke to clear, but then it came to me.
Someone had cold straightened the shaft material before making the shaft. Once the rotor ran and load and the laminations heated up, the rotor shaft took back it's bow.
Solution, new shaft.

That makes sense Ron. That could also be why it is insensitive to weight at the ends. It wants weight in the center.

That leaves the high slip speed. I agree with Ralph that that is very high if the motor is "SOLO" which I took as uncoupled.

Madhu, another possibility is "foot related resonance" at the NDE. This can cause very high vibration. What I would do is, with the motor running, loosen one - and only one - of the NDE motor feet and observe any changes. If no changes are observed, I'd retighten that foot. Then I'd loosen the other NDE foot. Then retighten. Then I'd do the DE motor foot bolts in the same manner.

I would never loosen more than one foot bolt at a time. A motor will run fine with 3 of 4 bolts tight. But this is only what I would do. I am not suggesting that you do this. Just sharing what I would do.

Also, would someone else run the numbers on the balance. The corrections I get are 3.9 gm @ +93 degs based on the trial. For the C1 run I get 1.9 gm @ +280 degs. For C2, 2.7 gm @ +288. For C3 3.8 @ +278. (The + means to move the weight X degs. relative to it's current position, wherever that is, because I entered the weight location as zero degs. for every run). The gram amounts seem awfully small, but possibly correct if running near a resonance. Another possibility is bad reference data. Balance "corrections" are only as good as the reference data. Is it possible that you changed something in your meter after the reference run? When a balance job doesn't go as expected, always take new reference data.

But remove the resonance, and the vibration will go away. This is a "foot related resonance" problem. I'd bet on it.

This message has been edited. Last edited by: rustythevibeguy,

A lot of good suggestions all around. I have two more cents to add.

I would also like to understand like Ralph asked: What was the exact speed? Does the speed make sense? (is there a vfd).

Also is the horizontal or vertical? Did you tell us what kind of bearings? Roughly how many horsepower (just like to form a mental picture of the machine).

I have seen a slow speed vertical motor with high 1x, non-directional that dropped immediately upon powerdown. See thread "Vertical slow-speed motor - high nondirectional 1x drops immediately on powerdown"

http://maintenanceforums.com/eve/forums/a/tpc/f/3751089...401009632#3401009632

Also, this behavior was intermittent. The motor had high and low vibrations several times.

There were a number of problems found. The one that explains the intermittent behavior I believe is that the upper thrust hub of this vertical motor (upon which the radial tilting pads also ride) was 6 mils loose. If the rotor shifts within this clearance and then somehow locks in at the new position by static friction, it can become eccentric.

The sudden decrease upon powerdown occurred on my motor and I think it's pretty easy to explain. People are familiar to expect magnetic force is 2*LF for static eccentricity and 1X+/-Fp for dynamic eccentricity (just like the wall chart). It is true for 2pole and to a lesser extent 4 pole / 6 pole. But it is not the case for slow speed motors such as my case. Simple theory tells you that the magnetic force for a slow speed motor will be constant (zero frequency) for static eccentricity and 1x for dynamic eccentricity. That is because the arc subtended by a pole is so small that poles are so close together that the variation in radial magnetic force among adjacent poles cancels out . What does not cancel out is that all poles near the small part of the gap have higher force. So it follows for slow speed motors: 1x for dynamic eccentricity, static force for static eccentricity.

Now, the fit of my scenario needs to be modified to describe Madhu's motor. For his motor with dynamic eccentricity we would expect 1x +/- Fp under load. But running uncoupled the Fp is virtually 0, so you can't see it (it may also show up as a very very slow increase/decrease in apparent level of the 1x, which could be another reason why dynamic eccentricity could appear intermittent while tested uncoupled).


I realize I'm starting to ramble so let me summarize with a few bullets:
- I think the sudden disappearance of 1x upon powerdown is an indication that the 1x is being "helped" by the magnetic field. This could be dynamic eccentricity from a number of sources, including excess clearance, rotor bow (as Ron suggested), simple eccentric rotor construction, etc.
- Dynamic eccentricity will show up as simple 1x while uncoupled because the Fp sidebands are too close to detect.
- Dynamic eccentricity may appear intermittent while uncoupled because the modulated 1x will come and go very slowly.
- Looseness of rotating parts that play a role in centering the rotor (such as our thrust hub) can also contribute to intermittent nature of the vibration dynamic eccentricity.
- An afterthought - rotor bow from loose bars can sometimes be intermittent.

This message has been edited. Last edited by: electricpete,