Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
mortenhuse

Member
Registered:
Posts: 13
Reply with quote  #1 
Hi all,

We have recently encountered what appears to be a winding failure in a motor of a set of large axial fan for tunnel ventilation. The specific motor is not dismounted yet so we don't know 100%, but the expected winding problem is based on uneven current measured on the different phases (45-90-87 A, nom 102 A @ 400 V 50 Hz) and that the motor protection triggers and stops the fan. The voltage of the three phases, measured in the mother board is all approx 400 V as expected.

We have recently started the service contract for this customer, but we will routinely every 6 month be doing vibration measurements on this and other similar fans, and specifically on the fan casing (very difficult to access the motor for direct measurements inside the fan casing). Direct measurements on motor will be done in case we find signs of failure or things we need to look more closely at.

My question is if there is any signs in the vibration spectra we have recorded (6 months prior to failure) which we have overseen? We would very much like to give the customer a heads up as early as possible for these types of failures and others (e.g. bearing, unbalance etc.), which we were not able to this time.

Please see the spectra attached. The measurements are done as 6-pack (2-10 000Hz) measurements in 3 directions measured on the fan casing. The name of the fan with winding failure is JV55. The other spectra are from 3 other identical fans placed adjacent to the one with failure. All fans are mounted vertically on spring suspension system. Last file (0942_001.pdf) is a sketch of the whole system.

Hope that some of you experienced vibration analysts could help us and improve our skills. All comments and tips are welcome!

Regards
Morten

Attached Images
jpeg Spectrum JV55 axial-vert_20160725.JPG (151.81 KB, 38 views)
jpeg Spectrum JV55 Rad-hor_20160725.JPG (162.31 KB, 38 views)
jpeg Spectrum JV55 Rad-hor-perp_20160725.JPG (139.95 KB, 33 views)
jpeg Spectrum JV55-JV58 axial-vert_20160725.JPG (141.61 KB, 31 views)
jpeg Spectrum JV55-JV58 Rad-hor.JPG (175.86 KB, 31 views)
jpeg Spectrum JV55-JV58 Rad-hor-perp_20160725.JPG (149.51 KB, 28 views)

 
Attached Files
pdf 0942_001.pdf (159.72 KB, 24 views)

electricpete

Sr. Member
Registered:
Posts: 652
Reply with quote  #2 

I don't ordinarily expect winding problems to show on vibration.  But if it did show, I would look for 2*LF (for example current unbalance might be caused by shorted turn or open parallel… either might also cause 2*LF if motor continued to run).  I think rbpf pattern might also be affected, but that tends to be more variable from many factors including loading and I don't pay attention to it. .  Nothing in your results would raise my suspicion of electrical problem personally.

Also could perhaps be high resistance circuit downstream of the location where the voltage is sensed.  Since it is a 2-speed motor, there may be some extra contactors in series with more possibility to introduce high resistance. I would do some more troubleshooting to narrow down whether or not motor is at fault before pulling the motor.  Electrical testing directly at motor terminals (winding resistance, insulation resistance, possibly surge).  Voltage checks at motor terminals (not remote location) while running.  With three phase motors we often rotate the leads directly at the motor to see if the oddball-current sticks with the same motor lead (motor problem) or the same power lead (supply problem) although this may be trickier with 2-speed motor. It may also be interesting to know how the current balance changes on the other speed.   

mortenhuse

Member
Registered:
Posts: 13
Reply with quote  #3 
Hi Electricpete,

Thanks for the swift reply and the skill fulled advises. We could not find any peaks in the vibration spectra that showed signs from this possible fault. However we have not been able to get a clear view on the amplitude of the rbpf, from what is OK and what is not. Just by comparing the rbpf and the FL (2x LF) with other spectra from fans/motors that are running fine, we could not find any deviations and then signs of error for this specific fan.

We totally agree with the troubleshooting that you propose and we will do this to the best of our ability. There are several practical challenges, but we intend to do as much as possible before dismounting the whole fan (which is vertically placed above the driving lanes in a very busy urban tunnel).

Just for info, the picture attached shows the only access point to the motor, through an inspection lid in the fan casing. In addition the motor has direct coupling from the motor cable to the windings and connection board.

Thanks again!

PA140087.JPG 

Ron Brook

Member
Registered:
Posts: 35
Reply with quote  #4 
If you have the capability to add Emax testing to your contract, that would be your best bet for a better diagnosis of what is going on electrically with the fan motor.  Agree with EP.  You need to test the motor more thoroughly before pulling the motor.   Test the motor with a megger at the first point you can in the motor leads from the motor.  If that means going through the access door and opening the motor terminal box, so be it.  Much better to know that motor is guaranteed blown before pulling it all the way out and then finding it to be good.  Had a similar issue with a vertical exhaust fan, 2xlf sidebands around rbpf as well as unbalanced currents at the VFD.  Opened the motor terminal box and found a hot, loose connection.  Redid the connection.....fixed!
Regards,
Ron Brook
Ginney

Member
Registered:
Posts: 2
Reply with quote  #5 
Good afternoon Gents, 

Is there not possible misalignment here based on the spike at 2 x fundamental?
Walt Strong

Sr. Member
Registered:
Posts: 889
Reply with quote  #6 
"Is there not possible misalignment here based on the spike at 2 x fundamental?"

How can there be misalignment (between 2 shafts) when fan rotor is mounted on motor shaft?

Walt
mortenhuse

Member
Registered:
Posts: 13
Reply with quote  #7 
Hi Gents,

Thanks for all the input and help guys. We will do further measurements and analysis at the next tunnel closing in accordance with the tips given here. We do not have the date yet, but hopefully quite soon. I will post our measurements/analysis then and hopefully we will find the cause of the problem. Thanks and "I'll be back" :-)

Morten
Allen Plymon

Member
Registered:
Posts: 29
Reply with quote  #8 
mortenhuse,

You have a lot of good advice already.  I was just looking at the speed of the machine that is marked at 933 CPM.  A 6 pole motor operating at 50 Hz would produce a nominal speed of 1000 CPM.  That is 67 CPM of slip.  Perhaps this is within the nameplate speed tolerance; however, seems like a lot.  The variance in amperage that you mentioned will certainly end a motor winding in a hurry.  If using only your existing monitoring package and adding more technologies is not currently an option, perhaps you could enter each motor current phase (current clamp meter) with a simple keypad input into your data collector for trending and phase to phase comparison.  And, add motor temperature that could easily be obtained with a pocket sized temperature gun - extremely advantageous.  Because with this amount of current imbalance, the motor temperature should have been excessive as this is what eventually caused the winding to fail.  Just thinking about how to use a potentially current program and add simple measurements to potentially catch this type of failure.

Best of luck!
electricpete

Sr. Member
Registered:
Posts: 652
Reply with quote  #9 
I agree 933 rpm would indicate excessive slip.
But I didn't see where that was labeled like that... most of the 1x labels I saw were up around 993 rpm.

Even 993 rpm would be a little low considering the drawing shows 995rpm full load speed, but 993rpm seems within a reasonable margin of error (not knowing actual line frequency, lines of resolution, and whether frequency interpolation was used to identify peak).
Allen Plymon

Member
Registered:
Posts: 29
Reply with quote  #10 
electricpete,

You are correct and I apologize for any confusion.  I'm not sure where I came up with 933.  But, glad that you caught that and corrected my error.  All the best!
mortenhuse

Member
Registered:
Posts: 13
Reply with quote  #11 
Hi guys,

Thanks for all the interest and help in our challenge. One of my technicians were in the tunnel doing some investigations this night and thinks he might have found the cause of the problem. In one of the connection boxes (connecting motor cables to power cables) the leads were slightly loose and in especially one of the leads it was clear signs of high temperature. The metal was melted and with clear signs of corrosion, expecting bad contact and high resistance in the conductor joint. Please see the attached photo. We do believe that this is the source of error, and we will change the connection box and test the fan afterwords.

Any thoughts gents?

Thanks again for all the help!

Morten

P5290047_ed.jpg 

Beatnik

Sr. Member / Supporter
Registered:
Posts: 183
Reply with quote  #12 
I don't think you will be able to see much while taking the vibrations datas on the fan casing. I'd try to have accelerometers and cables installed on the motor bearings during shutdowns. It would cost some money but your work would be much more efficient.
electricpete

Sr. Member
Registered:
Posts: 652
Reply with quote  #13 

Mortenhuse -  Thanks for feeding back your findings.

I agree with you -  that is very likely the smoking (no pun intended) gun that explains the current unbalance.  After fixing the connections, I would fire it back up (same motor) and measure the current balance as final post-maintenance test.

Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,403
Reply with quote  #14 
I'm not going to be of any help with the motor issue but I have found one case of high 1 x running speed that was imbalance due to what was described to me as a "hole blown in the rotor winding".  

I never saw it but I was asked to look at a small boiler ID fan that my client thought was vibrating.  There was high 1 x rs and a significant impact in the twf. We stethescoped our way around the motor with an ultrasonic device and found a circle around the stator where it was really loud. The motor was changed and the "hole" in the rotor winding was found.

This motor was only about 150 hp NEMA B regular old motor, though.
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.