Electrical Motor NDE high vibration

Dear all,

I'm monitoring the vibration of a 1.2MW vertical submerged pump(RPM: 595/ 9.917 Hz ). The motor 1x radial vibration appears to be high in one direction (pump discharge side). Radial vibration at 90 deg seems relatively low. I have checked the alignment at the coupling and any structural looseness and all seems ok.  

The Motor acceleration spectrum shows harmonics of 734 Hz with a 2xline frequency side-bands. I'm assuming this is rotor bar or a winding slot pass frequency. I'm not sure of the # of rotor bars or winding slots. 

I'm wondering if the high 1x vibration in possibly caused by the RBPF/WSPF??  

The coupling bolts were loosened and re torqued on 22/11/18 which showed some improvement temporarily(1x reduced from 18mm/s to 8mm/s). but the vibration levels increased within hours of operation. I have attached acceleration and radial velocity spectra in the discharge direction (marked R1) and at 90 deg to R1 (marked R2). In the R1 spectra, the green plot shows vibration immediately after coupling adjustment, and blue plot, a few hours later.   

Is this cause for concern?? Can anyone help with this??

Kind regards,

Pramukha  

Attachments

Original Post

About acceleration: It is very likely you have to deal with RBPS. To confirm RBPF you should have all the data to calculate it.  Do you have any measurement history? It is not necessary the amplitudes are too high for this machine. Some high resolution spectra should help you to see the exact matching with the multiple of rotation frequency. It could also be the change in load between your two measurements.

Velocity: The situation described shows high possibility of resonance present. You could do some bump tests in the critical direction to confirm the resonance. Thereafter you have some effective solutions to avoid the resonance. However you said the vibration level jumped back from 8 mm/s to 18 mm/s within hours but from the data you have sent, the max amplitude is still 10,6 mm/s .  

This to me looks more of a resonance issue rater RBPF. please give more details

  • How old is this equipment setup? I mean the motor & pump.
  • what is the trend of vibrations? are they high since commissioning or its a recent increase.

If you are concerned about RBPF & do not have the number of rotor bars, you can still go for motor current signature analysis.

Thanks

Akhil

Thank you both for your comments.

Here's a bit of background about the setup. Pump was commissioned back in 2008. The pump was overhauled back in 2013 due to pump failure due to bearing defect and again in 09/2018 (suspected from vibration analysis which turned out to be the case in 2018). This 1x vibration increase only occurred since the September bearing change. The condition monitoring program has only been going since August 2018 so not much data to go with. We do however have a twin pump and the 1x vibration at the same point is about 4 mm/s. The 734 Hz and 934 Hz peaks are also about half of this pump at the twin pump.  

As for points raised by Becar;  Correct! the vibration did not go back up to previous level of 18mm/s but it did increase to 10.6 within about 3 hours. My apologies! I realized after your comment that my statement was misleading.  

There were no change to load conditions between readings. 

I would like some clarity on why you both think this is resonance please. 

Regarding a bump test; My analyser was purchased back in 2011 and not what you would call cutting edge(FAG detector). so it takes about 1 min to complete measurement of 3 time signals of (3+3+0.3) secs at 3200 lines resolution(max). I'm not  sure how I can do a bump test with this. 

Next step was to do a motor current signature analysis. But we need to hire or potentially purchase the equipment for this. I'm wondering if it would be worthwhile in this case.      

  

 

 

 

 

 

The reason to suspect resonance is high vibration level in just one direction and changing amplitude over time. This is from my experience with similar vertical pumps. I suspect the construction stiffness is lower in one direction and could be strengthened with additional welding supports at the bottom - at the base plate. However prior to this I would do field balancing to minimize the 1x and see the difference.

Thimba posted:

Thank you both for your comments.

This 1x vibration increase only occurred since the September bearing change. The condition monitoring program has only been going since August 2018 so not much data to go with. 

But sounds like you are confident that judging from the data you have (in August), the problem was not there prior to the bearing change in September. Is that correct?

I would like some clarity on why you both think this is resonance please. 

I do not know why anyone else thinks there is a resonance problem, but it is one guess worth testing out. In my opinion without confirmation, resonance is only one possibility of many possibilities, hence, the need to confirm if it is or if it is not the problem.

Regarding a bump test; My analyser was purchased back in 2011 and not what you would call cutting edge(FAG detector). so it takes about 1 min to complete measurement of 3 time signals of (3+3+0.3) secs at 3200 lines resolution(max). I'm not  sure how I can do a bump test with this.

You should be able to do a bump test with your meter if it has a "monitoring function" for just monitoring data live.

Next step was to do a motor current signature analysis. But we need to hire or potentially purchase the equipment for this. I'm wondering if it would be worthwhile in this case. 

Can you answer the question in the attached plot of your data?  You may have to take data at a higher resolution to measure them, if they are even sidebands at all.

Thanks and Have a Great Day,

Ralph

Pramukha Plot

  

 

 

 

 

 

 

Attachments

Photos (1)

Hi Thimba

  1. Vertical orientation of machine - Any issues with pump also are reflected at Motor non drive end or cooling fan side or outboard side with possibly 1x running speed. but you recently changed pump bearings
  2. Pump is submerged - If pump vibration data can be taken it would be good to conclude, however l doubt the possibility for submerged case.
  3. The condition monitoring program has only been going since August 2018 - quite possible for the failures before the same indications were given by the equipment i.e high vibration with 1x at motor NDE
  4. When you loose the bolts the vibration changes - Very common thing in vertical machines as you are playing with the equipment stiffness.
  5. Rotor bars - I would ignore Rotor bar frequency if I don't have a trend of it. I see rotor bar frequencies in many motors & then keep a watch on its amplitude & the side bands.

You recently changed the bearing after which the vibration increased there is something in this activity that is the cause.

  • What about the motor junction box direction? Hope it has not been altered.
  • What is the pump bearing number? Anti friction bearing?
  • What was the vibration before pump overhauling?
  • Motor foundation bolt tightening which brought down the vibration can be re attempted with different pattern.

Thanks

Akhil

This sounds familiar to other Vertical Pump issues.  I have experienced elevated 1x pump/motor speed vibration  in the direction of the discharge of vertical pumps. Often the discharge flow from the pump will influence the Upper Motor bearing vibes due to the pipe not supported properly. The pipe flexing up/down pulls on the motor causing the increased 1x. I now will take phase/amplitude readings along the top of the pipe to at lease the 2nd or 3rd support away from the motor. This can be caused by minor flow changes, fluid temperature changes, and one case in Tampa the pipe temp was affected by afternoon sun light causing a dramatic temperature change on the pipe increasing the vibration levels on the motor. 

As other have suggested additional data needs to be collected at the foundation, motor to foundation interface, and the discharge piping. 

G. Kaiser

Add Reply

×
×
×
×