Dear Analysts,

We are facing an issue on our cement plant. It's raw mill main drive motor of 4300kw. On no load condition, vibration is normal but as the feed come, vibration reaches to 4-6 mm/sec (fluctuating between 4 and 6) and whole vibration is on 85Hz (which is natural frequency of end shield bearing housing as told by manufacturer). According to manufacturer, this frequency is being excited by some forcing frequency which is unknown.

Second thing which is observed is vibration amplitude is greatly influenced by line frequency. Vibration exceed 4mm/s as line frequency exceeds 50 hz, and it reduces to 2mm/s as line frequency drops below 50 hz i.e 49.9 hz.

Thirdly, at no load, dominant frequency is 2Lf (0.3mm/s) whereas at load conditions 2Lf is there (same magnitude i.e. 0.3mm/s) but dominant frequency is 85Hz as discussed above.

Fourth observation is, 2Lf and 85hz disappears as the motor is switched OFF.

Spectrums are attached for reference. Please guide.

PS: An year ago, insulation resistance of this motor was dropped below allowed value and then it was sent to vendor where is rotor was cleaned and painted/polished.

Attachments

Original Post
John from PA posted:

Hmmm, load dependent.  When is the last time you did a good look at the coupling?  Alignment, lubrication, wear, etc.

Thanks for reply. Coupling pin bushes are replaced with new, bearings are also replaced along with bearing housing. Alignment verified and found no issue of misalignment.

Yes BECAR, this is recent vibration change and it all started after activity mentioned in PS of initial post. Before it was running smoothly with vibration no more than 3. We want to dig for the cause of this vibration and this strange behavior. Specially its relation with Lf. Since it's 6 pole motor, if we consider maximum fluctuation in Lf, it brings 10rpm speed change at maximum. Vibration magnitude changes drastically with this difference.

One other thing is, at no load vibration magnitude is not influenced by Lf. This change in magnitude with change in Lf is observed only at load.

Dear Becar, motor is coupled with a single stage gearbox (GMF~640hz) which then drives the sun of a planetary gearbox box. Details of planetary gearbox are unknown and we've requested the supplier to provide the details. Sun of planetary gearbox runs at 40rpm. If dun has even 50 teeth then it's GMF will be around 33 Hz which is far less than 85 Hz.

We've collected vibration on gearbox body and it's noticed that two peaks of around 84.5 Hz and 86.9Hz with almost same magnitude i.e. ~1 mm/s.

But before motor maintenance that I mentioned in the starting post, there was no such issue. 2Lf is also visible in the vibration which tilts is toward motor problem.

Dear Becar, motor rated rpm is 995. We noticed from the display of control room where values of Lf and vibration values from installed vibration sensors are coming that fluctuation of Lf (which is almost between 49.8 and 50.2 Hz) causes  change of vibration from 2mm/s to 4mm/s or even more. When Lf is 50.2Hz, vibration will be 4 or 5 mm/s and when Lf drops below 50hz, vibration values decreases to around 2mm/s. Maximum fluctuation in Lf is about 0.5Hz.

Speed(RPM) = 120 f(Hz) / P(Poles)
So Fluctuation in speed will be 120*0.5/6=10RPM

So we can say that maximum fluctuation in RPM due to fluctuation in Lf is 10rpm causes change in vibration from 2mm/s to 4mm/s.

Abid.Khan posted:

Dear Becar, motor rated rpm is 995. We noticed from the display of control room where values of Lf and vibration values from installed vibration sensors are coming that fluctuation of Lf (which is almost between 49.8 and 50.2 Hz) causes  change of vibration from 2mm/s to 4mm/s or even more. When Lf is 50.2Hz, vibration will be 4 or 5 mm/s and when Lf drops below 50hz, vibration values decreases to around 2mm/s. Maximum fluctuation in Lf is about 0.5Hz.

Speed(RPM) = 120 f(Hz) / P(Poles)
So Fluctuation in speed will be 120*0.5/6=10RPM

So we can say that maximum fluctuation in RPM due to fluctuation in Lf is 10rpm causes change in vibration from 2mm/s to 4mm/s.

Thus phenomenon is quite strange for us and it only in load condition. This fluctuation is not observed in no load condition.

I wouldn't say your situation is strange when you wouldn't say it was not present before. Speed and vibration fluctuations you are witnessing are nothing special at machines with changing load. And raw cement mills are definitelly in that group. When you are 100% sure it is a new phenomena, I would bet it is related to some motor's drop of power after overhaul and could be related to electric motor issues. I would also check if anything changed on the electric network this machine is connected to. Enough kV for the motor? 

We'll check if anything changed on the electric network this machine is connected to.

We also acquire current and flux data. This data you can see that data in attached file (file that is attached to starting post).

Can there be any source of 85Hz in motor? Or if the source of 85 Hz is planetary gearbox, can a frequency from that far location excites motor outboard housing natural frequency? Motor and gearbox are coupled by flexible coupling and installed on separate foundations. And what further test do suggest for motor rotor condition?

Or if the source of 85 Hz is planetary gearbox, can a frequency from that far location excites motor outboard housing natural frequency?

Sure, it can. But there also comes the same question, why it appeared after motor overhaul.

What type of bearings are there? I suppose fluid type?

Motor and gearbox are coupled by flexible coupling and installed on separate foundations.

I don't like the fact about the separate foundation. It also could be causing some problems.

 

I must say I suspect two other options rather then bearing frequency:

- something is exciting your motor from the side of the gearbox / planetary gearbox, 50 teeth for sun gear sounds a low number, you also have to check the planetary teeth rotation speed

- or there is some electric fault weakness on the motor but I don't understand how is it possible that the line frequency is fluctuating?? OK, rotation frequency, but line frequency should be constant at 50 Hz 

But it could only appear when there is a fault on the balls. Secondly, it's higher harmonics appear first and as the fault grows, ores 3rd, 2nd and 1st start to appear. Even though be bearings were installed, I checked the spectrum, there was no any other other frequency. I will reconfirm on Monday as well.

Becar posted:

I must say I suspect two other options rather then bearing frequency:

- something is exciting your motor from the side of the gearbox / planetary gearbox, 50 teeth for sun gear sounds a low number, you also have to check the planetary teeth rotation speed

- or there is some electric fault weakness on the motor but I don't understand how is it possible that the line frequency is fluctuating?? OK, rotation frequency, but line frequency should be constant at 50 Hz 

We are waiting for reply of supplier regarding details of planetary gearbox. As I mentioned in my earlier post that we have also seen two peaks i.e. 84.5 and 86.9 on the gearbox body. I'm sharing the spectrums here as well. I think there is a allowed range in which Lf can vary. Due to this variation, rpm variate.

Attachments

Rpm variates because of the variable load not because of the variable line frequency. It is similar when you drive a car from the city to the hills. You can have your gas pedal at constant but your speed changes. I am not an electrician expert, but I think the line frequency should be constant because it doesn't come from the machine but from the distribution network. Would wait for some other electric expert to tell the truth  But the motor should be catching the line frequency not the other way.

It’s tough to piece it together.  First glance variation with frequency suggests resonance excited by something that varies with running speed, but what? (the $64,000 question).

You mentioned OEM suspected endshield resonance. I have seen that on large motors. Typically the end of the motor has a big plate or inspection cover on the inboard and outboard end, something like a semi-circle above the bearing.  Typically it vibrates with a drum-head type resonance and it can easily be verified by bracing (or sometimes just touching) the center of that piece… which (in our case) brings the vibration down all around the machine.   Your geometry may be a little different, in that case try mapping the axial vibration around on the end surface of the motor and try temporarily axially bracing the place with the highest vibration.  If you confirm that type of resonance, the fix may be as simple as providing some stiffening to that piece.

Becar posted:

- something is exciting your motor from the side of the gearbox / planetary gearbox, 50 teeth for sun gear sounds a low number, you also have to check the planetary teeth rotation speed

- or there is some electric fault weakness on the motor but I don't understand how is it possible that the line frequency is fluctuating?? OK, rotation frequency, but line frequency should be constant at 50 Hz 

Actually, in my experience 50 on the sun would be quite high for a raw mill application.  In general when shock loads are high, then the pitch gets "course" which lowers the number.  

As a reference to a typical drive and the associated vibration analysis, see http://www.vibration.org/Prese...0Sept%205%202006.pdf.

Is this a multiple stage gearbox and do we know (for sure) is it a planetary arrangement (as opposed to a sun or solar arrangement).

I apologize, I was thinking about ring gear not sun gear. My mistake. Wouldn't be logic for sun gear!

I am still very interested in this case specially about line frequency oscillation and I hope some expert tells the truth. Is it possible in this case? I know el. power can change, motor current.... But line frequency??

John from PA posted:

Is this a multiple stage gearbox and do we know (for sure) is it a planetary arrangement (as opposed to a sun or solar arrangement).

There are two stages of this gearbox; first is simple gearbox with bevel gear and is driven by motor. Output shaft of this gearbox run at 40rpm which drives the sun of second stage of gearbox which is a planetary (epicyclic) gearbox. Carrier of this gearbox drives the table of Raw Mill.

Abid.Khan posted:
John from PA posted:

Is this a multiple stage gearbox and do we know (for sure) is it a planetary arrangement (as opposed to a sun or solar arrangement).

There are two stages of this gearbox; first is simple gearbox with bevel gear and is driven by motor. Output shaft of this gearbox run at 40rpm which drives the sun of second stage of gearbox which is a planetary (epicyclic) gearbox. Carrier of this gearbox drives the table of Raw Mill.

Your motor speed is stated as 995 RPM and the input to the planetary is 40 RPM yet you call the first gearbox single stage.  That would be almost a 25:1 ratio, a bit unusual for a bevel.

Is this a Maag vertical mill drive?

John from PA posted:
Your motor speed is stated as 995 RPM and the input to the planetary is 40 RPM yet you call the first gearbox single stage.  That would be almost a 25:1 ratio, a bit unusual for a bevel.

Is this a Maag vertical mill drive?

Sorry, my mistake, 40 rpm is the speed of last shaft i.e. Carrier of planetary (raw mill table). I just received death details of bevel gear gearbox. Number of teeth in input gear are 14 and number of teeth on output gear are 52. Hence rpm of output bevel gear shaft (and the sun) are 268.

And yes, motor is horizontal and raw mill is vertical.

Abid.Khan posted:
Sorry, my mistake, 40 rpm is the speed of last shaft i.e. Carrier of planetary (raw mill table). I just received death details of bevel gear gearbox. Number of teeth in input gear are 14 and number of teeth on output gear are 52. Hence rpm of output bevel gear shaft (and the sun) are 268.

And yes, motor is horizontal and raw mill is vertical.

How much else in the way of pertinent info may be in error?

By the way, in your calculation (see below) the "120" is line frequency so unless you have a 60 Hz source, the result of the calculation should be 8.33 RPM.  But your motor speed of 995 agrees with a 6-pole motor on a 50 Hz supply.

Speed(RPM) = 120 f(Hz) / P(Poles)
So Fluctuation in speed will be 120*0.5/6=10 RPM

 

Akhil Rathore posted:

Hi Abid

Are there vibration sensors installed on the gearbox. Are you capturing data online?

 

Image that I shared in which fluctuation of vibration values of three sensors with respect to fluctuation in line frequency are permanently installed sensors. However the pdf file and .rtf file that I shared are captured wit portable analyzer CSI 2140.

Instrumentation cables are same.

John from PA posted:
How much else in the way of pertinent info may be in error?

By the way, in your calculation (see below) the "120" is line frequency so unless you have a 60 Hz source, the result of the calculation should be 8.33 RPM.  But your motor speed of 995 agrees with a 6-pole motor on a 50 Hz supply.

Speed(RPM) = 120 f(Hz) / P(Poles)
So Fluctuation in speed will be 120*0.5/6=10 RPM

Dear Sir,

That mistake was due to some miss-understanding. There will be no other error in the provided information.

I think 120 is a constant in the formula (not Lf) whereas f(hz) is the line frequency. So to calculate the fluctuation of RPM, I used maximum fluctuation of frequency.

ac-motor-synchronous-speed

Attachments

Photos (1)

I think you should trace the cables as I experienced a similar issue where a gearbox was having frequencies that I was not able to correlate with any defect frequency or GMF and the vibration was fluctuating wrt the power freq. The issue was since commissioning of the plant.

Akhil Rathore posted:

I think you should trace the cables as I experienced a similar issue where a gearbox was having frequencies that I was not able to correlate with any defect frequency or GMF and the vibration was fluctuating wrt the power freq. The issue was since commissioning of the plant.

Thanks for suggestion. But same frequencies were observed in the vibration data that was captured by portable analyzer. Values of vibration were also fluctuating when I collected the vibration data in monitor mode in 2140 analyzer. That time, while collecting offline data, we could not see what the line frequency was. We'll do the trading of cables too to rule out this possibility.

Same was in my case 

the data I was collecting using portable analyser and the casing mounted sensors both were showing similar “ghost peaks” and were fluctuating with power frequency.

But in your case the issue has started after a maintenance activity. 

I would suggest you to see a waterfall plot and locate when this peak has started appearing. Was it present before the maintenance activity or not. Then list out ALL the activities you did during the maintenance and then strike one by one. 

Hope you have checked for bearing fault frequencies

confirmed that this peak is not from the driven component through motor solo

 

Akhil Rathore posted:

 

Hope you have checked for bearing fault frequencies

confirmed that this peak is not from the driven component through motor solo.

Bearings along with new bearing housings have been installed. On gearbox body, two close peaks i.e. 84.5Hz and 86.9 Hz were observed each with magnitude ~1mm/s. Spectrums have been shared in my earlier post (can share again if required). However on motor it's single sharp peak. 85Hz peak is also visible in motor solo run but of smaller magnitude. At no load as well as on solo, vibration magnitude are quite low.

If its in motor solo run also then you know it is not related to the driven equipment 

you also said as you cut off the supply the peak vanishes

verifying existing cabe laying is not easy but you can still check if some new cables were laid during the shutdown and were they laid close enough to the sensor cables 

 

just my opinion I could be wrong 

> 85Hz peak is also visible in motor solo run but of smaller magnitude. 

I think this tends to confirm there is a resonance in the motor suggested by the OEM. When coupled it is being excited by a gearbox frequency (that excitation shifts toward/away from the resonance as frequency changes, explaining the frequency dependence). When uncoupled, it is excited by the small amount of broadband excitation available at the motor.

Rhetorical question: What else could possibly cause a frequency like this in a motor solo?? I don’t think it’s bearings, because:

  • 1 – bearing defect wouldn’t be frequency-sensitive unless coupled with a resonance
    2 – you told us you don’t have harmonics or other related frequencies
    3 – you changed the bearings

I can’t think of anything else. (It’s a resonance imo)

It’s often easy enough to check (and potentially fix) this endshield resonance. Map the axial magnitude of the offending frequency on the external ends of the motor frame. If you find the vibrations at some location (usually near the center of a large sheet-like structure) are higher than the bearing housing, that’s your confirmation. Try bracing it where magnitude is highest (the anti-node). If it’s like the one I saw, vibration across the machine will go down when you brace that spot, maybe even just when you touch it lightly. Or if you get stuck post some photos or vib measurements. It’s easy and worth a try imo.

Dear @Akhil Rathore, yes 85 Hz is present in the solo run. We will verify cables layout. What should we check in verifying cable layout other than laying of new cables in vicinity of sensors cables?

Dear @electricpete, What else could possibly cause a frequency like this in a motor solo?? ($64,000 question). Vibration in axial direction is low as compared to radial direction (peak at 85Hz still there). I'm attaching vibration data in velocity for reference. However, there is a small motor driven fan (other than motor rotor cooling fan which is installed at the end of rotor) installed on the body of motor with metallic duct to cool carbon rings/brushes by circulating air. Horizontal direction on this small motor is axial direction of the motor under discussion. When vibration on main motor is 3, vibration on this small motor and associated ducts is 7 and on the same dominant frequency. We changed this motor and fan with similar motor & fan to rule out of this motor is source of problem but vibration remained the same. Now we are planning to remove this motor completely (and keeping circulation of air by other means) to recheck its effect on main motor vibration. But we can do this only when having sufficient raw material stored so that stoppage in production may not occur.

Attachments

Hi Abid

You just have to see that any electrical cables are present in vicinity of vibration sensor cables or are they crossing each others path at a point.

I doubt removing the small motor will change any vibration but would still be interested in the results as @electricpete was doubting resonance and you are removing some mass from the system, so that may ...i dont know.

 

 

Akhil Rathore posted:

Hi Abid

You just have to see that any electrical cables are present in vicinity of vibration sensor cables or are they crossing each others path at a point.

I doubt removing the small motor will change any vibration but would still be interested in the results as @electricpete was doubting resonance and you are removing some mass from the system, so that may ...i dont know.

I'll post the result as soon as any change occur.

I agree it's not as likely endshield resonance if it's not showing up axially. You changed the housings - was the purpose of that action to address a suspected resonance, or something else?

> there is a small motor driven fan (other than motor rotor cooling fan which is installed at the end of rotor) installed on the body of motor with metallic duct to cool carbon rings/brushes by circulating air.

Brushes suggest (the big motor is) either wound rotor induction motor, sync motor, or dc motor.  

Slip rings narrow it down to wound rotor induction motor or sync motor (dc motor would have "commutator" which has insulated segments unlike slip rings which are continuous smooth). 

Do you know what type of motor it is?

Is there a drive or vfd for the motor? 

 

 

Hi Abid,

I got a motor vibration issue last year. It is also driving a gearbox and then a large roll. And I found the high vibration is caused by distorted harmonic provided by VFD.

Could you please take some additional measurements:

(1) Can you check PeakVue at both DE and NDE with a 2K HP?

(2) Can you take a long measurement (maybe 12800 points) with Fmax around 2500 Hz, and show the low frequency and high frequency range (roughly around RBP)?

(3) When motor is not loaded, take measurements on four motor feet in vertical direction, and provide 1X and 2X amplitudes. If possible, conduct a simple ODS.

(4) It will be much easier if you can get instantaneous rotational speed (measured from tacho) from control guys.

Abid.Khan posted:

Dear Analysts,

We are facing an issue on our cement plant. It's raw mill main drive motor of 4300kw. On no load condition, vibration is normal but as the feed come, vibration reaches to 4-6 mm/sec (fluctuating between 4 and 6) and whole vibration is on 85Hz (which is natural frequency of end shield bearing housing as told by manufacturer). According to manufacturer, this frequency is being excited by some forcing frequency which is unknown.

Second thing which is observed is vibration amplitude is greatly influenced by line frequency. Vibration exceed 4mm/s as line frequency exceeds 50 hz, and it reduces to 2mm/s as line frequency drops below 50 hz i.e 49.9 hz.

Thirdly, at no load, dominant frequency is 2Lf (0.3mm/s) whereas at load conditions 2Lf is there (same magnitude i.e. 0.3mm/s) but dominant frequency is 85Hz as discussed above.

Fourth observation is, 2Lf and 85hz disappears as the motor is switched OFF.

Spectrums are attached for reference. Please guide.

PS: An year ago, insulation resistance of this motor was dropped below allowed value and then it was sent to vendor where is rotor was cleaned and painted/polished.

You have not marked any of the peaks? Is the peak of concern EXACTLY 85 Hz?

electricpete posted:

Brushes suggest (the big motor is) either wound rotor induction motor, sync motor, or dc motor.  

Slip rings narrow it down to wound rotor induction motor or sync motor (dc motor would have "commutator" which has insulated segments unlike slip rings which are continuous smooth). 

Do you know what type of motor it is?

Is there a drive or vfd for the motor?

Sorry for delayed response. Motor is a three-phase asynchronous slipring motor (VEM Sachsenwerk GmbH Type DSRAN 9034-6WF).

Vibration_Charlie posted:

Hi Abid,

I got a motor vibration issue last year. It is also driving a gearbox and then a large roll. And I found the high vibration is caused by distorted harmonic provided by VFD.

Could you please take some additional measurements:

(1) Can you check PeakVue at both DE and NDE with a 2K HP?

(2) Can you take a long measurement (maybe 12800 points) with Fmax around 2500 Hz, and show the low frequency and high frequency range (roughly around RBP)?

(3) When motor is not loaded, take measurements on four motor feet in vertical direction, and provide 1X and 2X amplitudes. If possible, conduct a simple ODS.

(4) It will be much easier if you can get instantaneous rotational speed (measured from tacho) from control guys.

Thank you very much for suggestions. We will perform these test as soon as possible depending upon process and its availability.

Viberman posted:
You have not marked any of the peaks? Is the peak of concern EXACTLY 85 Hz?

Peaks are not exactly at 85Hz and file with marked peaks is given in my earlier post which is linked here Click here.

Walt Strong posted:

The plots indicate that 85-Hz is not exact frequency, and it may actually be 5x shaft speed. Time synchronous averaging could confirm this or rule it out.

Walt

Yes, the peaks are not at exact 85Hz. But these are not 5x of shaft speed. These peaks are non synchronous.

Thank you for the link to the earlier plots. First, a couple of observations of what I think I see in the data.

1. I noticed that the motor outboard horizontal and vertical amplitudes at the nominal 85Hz are not all that different. It is very unusual for resonant frequencies to be the same horizontally as vertically... possible, but unusual.

2. The inboard vertical on the motor does not appear to be really involved, so the "amplification" (if it is amplification) is worst outboard and confined to the horizontal inboard. 

Got to go, but I will keep an eye on this string and contribute if I come up with anything. Happy hunting.

 

Add Reply

Post

HACKATHON in association with IMC-2019

×
×
×
×