Hello All,

Long time lurker, first time poster! I apologize in advance for the length of this post.

I'm having some challenges determining the root cause(s) of some vibration on an 800 Hp, 1200rpm, center hung, forced draft fan & motor for one of our clients. Both fan and motor have sleeve bearings and the fan is directly coupled via grid style coupling. The motor has been recently serviced after it had failed on site, and the fan bearings were also removed and confirmed to be in proper operating condition (or so I'm told). No proximity probes are installed in any of the bearing housings, so the only data I have is what I collected with standard accelerometers.

Yesterday, I was asked to collect vibration data as the fan was fired up and run through different operating loads. Motor speed is constant, and load is adjusted by opening and closing the dampers. What I observed was somewhat reasonable vibration levels, albeit a little higher than I'd ideally like, from 0% up to 75% load, predominantly 1X running speed along with very minor harmonics. However, once they hit 75% I started to notice visible, non-periodic axial pulsations that were quite evident right at the coupling, as both motor and fan shaft appeared to be shifting axially. My best guess for this issue at the moment is that the air flow becomes turbulent under this load condition and thus the fluctuating pressures cause the axial movement as the fan draws fresh air from the one side exclusively. 

After the axial movement was observed load was reduced down to 50% and I continued to monitor vibration for another 30 ish minutes as the equipment continued to come up to operating temperature. As things continued to warm up the vibration amplitudes continued to increase, predominantly on the motor. The final set of readings I collected showed amplitudes upwards of 0.5 in/s at 1X in the horizontal axis on the NDE of the motor. Phase readings across the face of the DE of the motor and fan bearing showed the phase being out by 60-90 degrees as I went around from 0 to 9 o'clock positions. With this information, and the developments in the spectral data (increasing harmonics and large increase in some 2X and 3X) I am inclined to believe that the thermal growth of the motor is causing misalignment, especially considering the fan bearings are liquid cooled and the motor bearings have no cooling whatsoever.

Now I understand that does seem like a heck of a lot of vibration from a bit of misalignment caused by thermal growth, but there is one more issue I believe to be contributing. As I was walking around the equipment collecting readings I noticed the vibration I felt where I was standing change drastically depending on where I stood. The motor and fan are supported by a raised concrete base, and this base is sitting on what I believe is further reinforced concrete, as there is a definitive cut line a few feet around the whole perimeter. As I stepped from inside to outside the cut line there was a notable reduction in vibration. When I measured vibration inside the cut line at the corner nearest the NDE of the motor I recorded 0.037 in/s peak amplitude at motor running speed. I then collected a reading just outside the cut line (~3 in away) and not only had the peak amplitude significantly reduced down to < 0.01in/s, the phase also lagged by approximately 60 degrees compared to just inside the cut line. This leads me to believe that the concrete foundation may be compromised (near the NDE of the motor) resulting in some structural instability, possibly as a result of the previous failure. I'm inclined to believe that this instability is allowing the vibration caused by the misalignment to run rampant, particularly in the horizontal axis.

I'm hoping to get some input in regards to the conclusions I've drawn regarding the axial movement, misalignment/thermal growth, and possible structural instability. Does it seem like I'm heading in the right direction? Or am I completely out to lunch/missing something completely obvious? I've included full spectral data (starting with 0% load, 20%, 50%, 75%, then back down to 50% for the last 3 readings as temperature increased), along with a couple pictures of the general layout of the equipment.

Any and all feedback is greatly appreciated!

Attachments

Photos (2)
Files (1)
Original Post

The first thing I would do is to find out why the shafts are moving axially seen from the naked eye because they shouldn't (are you sure that the shafts are moving or could it be the coupling hubs?). The harmonics could be saying you are dealing with some mechanical looseness. The second thing is to check the alignment in the hot condition after shutdown. Did you balance the fan rotor? Was the motor balanced while on service. I would not pay much attention on your concrete base findings at this moment because it is not unusual to have the vibration higher in the inner base.

Thanks for the reply, Becar.

I agree, the axial movement is definitely concerning, which I can say with a fair degree of certainty is in fact the shafts moving and not the coupling hubs. There were several of us that notice around the same time, and that's when the decision was made to drop back down to 50% load where it no longer occurred.

I can check with the maintenance staff on site, but I'm pretty sure every bolt and nut has been double and triple checked at this point. It might also be worth noting that the sleeve bearing clearances have all been checked as well, and if anything are on the tight end, but still within the desired range. They have also inspected inlets and outlets along with the rotor for any debris buildup, and have tested the rotor for any cracks or damage.

I had suggested they perform a hot alignment and I believe that is happening this afternoon. Hopefully will hear of some results soon.

I have not balanced the rotor, nor is there any available balancing history from our client. I know it can always be made better, but if it were a balance issue would it not affect the radial vibration amplitudes in a similar manner, and not be so greatly amplified in the horizontal axis?

On first reading, one might assume it was load induced. However, I suspect that to simply be coincidence as the machine got up to normal operating temperature....and where things start to go awry. I think this was proved by your statement "...load was reduced down to 50% and I continued to monitor vibration for another 30 ish minutes as the equipment continued to come up to operating temperature. As things continued to warm up the vibration amplitudes continued to increase, predominantly on the motor." I will assume that the plots you show are from "latest to earliest" as you go out along the Z (time) axis.  Yes?

Sounds like you've put a lot of effort into this so far and fairly well documented. Whenever something changes from previous, my immediate focus is what went on during the overhaul, and in your case it primarily involved the motor; hence I agree with you and Becar that accounting for thermal growth/precision alignment is the first potential culprit to evaluate (as it would explain many of the findings so far) and one that would be affected by motor replacement. Is the initial vibration data when first started similar to previous vibration levels prior to motor replacement?  I also agree with Becar that being too concerned about the concrete base differences are "nice to know but not likely" part of the problem at this time.

Also confirm with your mechanics that any required clearances between the motor and pump shafts are met. 

Hopefully your hot alignment check will confirm suspicions. Regardless, it is always a good idea to record vertical motor-frame temperatures from base to bearing to see if this supports this possibility.

Good luck and let us know results

Regards

Jim P

NOONY,

Waterfall plot shows what I believe to be weak fan base/support structure. KISS principle, motors typically do not vibrate, if you took the motor off and ran on the floor or test past it will most likely not vibrate from cold to hot conditions. If you put any motor/machine on a weak foundation you stand the chance to excite resonance. The additional load being placed on the machine with the damper can cause the base to twist like a V8 engine when you rev it up, it twists.

As you read this post it smells of load related movement and possibly imbalance. Directional vibration, vibration changes with load, the vertical does not change vibration levels with load only horizontal, when changing fan bearings it always a good idea to trim/balance the fan wheel.

Have you taken phase data on the base? Take vertical vibration and phase readings on the base closest to the concrete. take 10 readings per side and see what it shows you. Typically I call a base anchored when the vibration levels are below .01 in/sec overall vibration

Do you have images from different angles of this machine and its base? The image of the fan, I suspect the "grout" that is poured on top of the concrete block is cracked and loose. This in turn does not properly support the machine base. You can tighten all the bolts you want but it will not fix loose concrete or grout. The evidence is the "rusty" liquid that is squirting down the side of concrete block, yep right at the motor drive end.

Dave

Thank you everyone for the replies! I'm thoroughly impressed with promptness of the responses.

Jim,

Yes you are correct about the spectral plots. The newest are at the forefront, and go back towards the earliest readings.

Thanks, Jim. I try to make a habit of collecting too much data rather than too little, and I feel presenting as much in an organized fashion (but not too much) is the least I can do when reaching out to the community for help. The motor I know to be running quite well on it's own. The motor rewind shop I work out of completed the repair, and I collect and analyze vibration data on all outgoing motors. This one had blown in the slot and did a fair amount of stator lamination damage which required extensive iron work prior to being rewound. Uncoupled in our shop it did hunt for magnetic center slightly but nowhere near to the extent of axial movement observed when coupled on site. Unfortunately the hot alignment did not reveal any issues, and measurements were within .001" of the original cold alignment values, which I've been told were near perfection.

I do not have any vibration data prior to the motor overhaul, as I was only brought it afterwards. However, there is an identical (mirror setup) motor and fan right next to this problem one that I also took readings on for comparison when I did my initial data collection and analysis, post motor repair, a few weeks back. It did have it's own gremlins, but horizontal vibration and possible base issues were not among them. That second motor is currently out for repair so I am unable to collect more readings for comparison. 

I will check with the mechanics regarding shaft clearances.

Dave,

I agree, my initial suspicions lead me to believe base/support issues as well. I may have just been focusing on the wrong part of the base perhaps.

Are you suggesting the weak base may be resulting in some torsional resonance as load is increased or more of a directional resonance? And why might this condition worsen as operating temperature increases, but while load remains consistent. I only ask because in my experience (albeit limited) anytime I deal with equipment operating in a resonant condition it appears right away and doesn't so much develop over time. 

I certainly don't disagree that load is causing some of the issues. What I'm struggling to sort out is what change has occurred to cause the system to react this way under load. The history of this equipment that I've been provided with is questionable at best, so for all I know this may not even be a new problem. I can definitely suggest to attempt an in place balance, although I'm worried this won't fully resolve the vibration issues being experienced.

I may have some additional phase data but I will have to double check when I get back to the office. Thank you for the suggestion. I will have to collect the relevant phase data next time I am at site, although I am fairly certain that the amplitudes will be above 0.01 in/s given what was recorded on the footing inside the cut line further out.

I have a few more pictures back at the office but I will be sure to get a few more good ones next time I'm at site (likely this Friday). The liquid seen going down the side of the base may also be bearing oil. They had been severely overfilling the bearing cavities in recent history (proper level now maintained) and it was leaking everywhere as a result. Again, I will try to confirm this as soon as I can.

Shabir,

Agreed, there is definitely looseness. Determining where this looseness is originating is proving to be an issue.

Motor runs good solo when tested after the repair at our shop. Very low overall vibration levels, although minor amounts of hunting for magnetic center could be seen.

Have not checked for soft foot although I can ask if the mechanics have done this. Would we not expect to see elevated 2X LF if this were the case though?

All,

Thank you all again for the feedback! If I've missed answering any questions please let me know. More info to come...

Cheers,

Sean

Are you suggesting the weak base may be resulting in some torsional resonance as load is increased or more of a directional resonance? Yes

And why might this condition worsen as operating temperature increases, but while load remains consistent. Depending on operating temperature/fan housing/fan wheel all grow with higher temps. If temp is not changing, FD fan, then the weak base can be the source.

I only ask because in my experience (albeit limited) anytime I deal with equipment operating in a resonant condition it appears right away and doesn't so much develop over time. I agree with your comment, but I also see older machines bases/foundations deteriorate over time and slide right into resonance, even 100-200 rpm speed change on belt drive and vibes go up or down due to resonance. If you get your nose right next to the steel base to grout interface you might even see bubbles from the oil/loose base.

I certainly don't disagree that load is causing some of the issues. What I'm struggling to sort out is what change has occurred to cause the system to react this way under load. Another observation, the steel base, has grout holes with no grout. The motor weighs more than the base it is sitting on!!! Typically I see at 3-1 ratio of mass vs machine, meaning concrete/base supporting the machine has 3 times the mass of the machine itself. For power plants this is 10-1 or greater. If the steel base is not securely anchored/supported to the concrete block then the steel base is supporting the full torsional load. If anyone lifted the motor and it tugged upward on the base, if the motor hit the base hard during repair procedure, oil saturating the grout causing it to crumble, etc

The history of this equipment that I've been provided with is questionable at best, so for all I know this may not even be a new problem. I can definitely suggest to attempt an in place balance, although I'm worried this won't fully resolve the vibration issues being experienced. Sounds like you have a good handle on next steps, its free information, more data helps paint the picture clear or limits the amount of what can be wrong.

Have you viewed the data like the below waterfall? Using all bearing locations and directions compared to each other? This is a motor (1,2) and frame 4 end suction centrifugal pump (3,4), Note the motor has the most vibration with 1x still present on  the pump, vibration is directional, this particular problem is base related, installed in Jan19. This kinda gives you an idea what is moving, what direction its moving, which location has the movement

Dave

Attachments

Photos (1)

[quote]Both fan and motor have sleeve bearings and the fan is directly coupled via grid style coupling[/quote]

[quote] I started to notice visible, non-periodic axial pulsations that were quite evident right at the coupling, as both motor and fan shaft appeared to be shifting axially[/quote]

Typically if you have a sleeve bearing motor, then the driven machine will have thrust bearings of some type with relatively small axial clearance (~0.010”) and will also have a "limited endfloat coupling" which limits axial movement of motor shaft with respect to driven shaft to some narrow range (typically less than  1 / 4”).  I am doubting that you meet either of these...

 The fact that you are seeing the fan shaft moving axially makes me doubt whether you have a thrust bearing in the fan.  Please investigate what thrust bearing is included in the fan and what is the axial clearance of that thrust bearing. 

 Also, while I’m not positive, I don’t think a grid type coupling is suitable as a limited endfloat coupling.

In summary, what is your understanding of how the axial position of the shaft and motor are limited in this machine? 

Hello everyone!

Sorry for the delayed response, it's been a fairly busy week. 

UPDATE:

A couple things have happened since my last post. The client's site electrical engineer performed a PdMA test while operating at 60% load. At first glance he suspected there may be rotor bar issues, however; after sending the data directly to PdMA for a second opinion they didn't see anything that would suggest such an issue. I'm not entirely surprised on PdMA's analysis, as I would have expected to see signs in the vibration data if there were in fact rotor bar issues.

They then removed the motor from service and sent it to another shop in the city to be load tested. I have been told that once a load was applied it started vibrating excessively, similarly to what was experience on site. The shop suspects "mechanical looseness, possibly a bearing housing fit". I have still yet to hear if anything specific has been found.

Jim, 

I apologize, I still haven't gotten an answer regarding shaft clearances during installation.

Dave,

Thanks for the waterfall data! I haven't seen it laid out like this before, but it definitely helps to visualize whats's going on. I'll have to see if my software will allow me to display the data in a similar fashion.

As the motor was removed from service I wasn't able to collect any more phase data, however; the one phase reading regarding the base that I do have shows a 55 degree phase shift between the concrete/grouted base the motor is sitting on and the footing below.

I've also attached more pictures of the base and general equipment layout. I don't have a ton of experience with identifying compromised bases/grout, so I'll trust your judgement when you've get the chance to look at the pictures for yourself.

ELECTRICPETE,

I've been told there is a "thrust collar" on the inboard end of the fan. Axial clearance was set at 0.011" (tolerance 0.010"-0.015"). I'm still at a loss trying to understand how so much axial movement could be seen with this in place... 

What type of coupling would be appropriate in this scenario?

All,

Thanks once again for all the responses! I'm sorry that more info isn't available yet. It's only been trickling in from our client.

Attachments

Photos (12)

Grout looks suspicious, lots of "liquid" something leaking down the side of concrete block. Cable 

As things continued to warm up the vibration amplitudes continued to increase, predominantly on the motor. The final set of readings I collected showed amplitudes upwards of 0.5 in/s at 1X in the horizontal axis on the NDE of the motor. Doubtful motor would be sent to you out of balance...when you state "warm up" does this also mean load was increased or just warmed up with steady state load? If the motor was .5 in/s what was the vibration on both fan bearings?

Dave

Dave,

I agree the liquid is suspicious, however; I'm not sure how I would determine whether it is a result of the client overfilling the bearing housings with oil for some time, or a compromised base. Unfortunately I wasn't able to observe the motor in operation when I was there for my regular data collection route last Friday, as they had already removed the motor from service. I was hoping to potentially see some oil bubbles coming out between the grout as you had suggested previously.

Motor was serviced at the shop I work out of and was balanced during the initial repair. Vibration data collected prior to delivering to our client showed amplitudes at all recorded frequencies < 0.01 in/s while under no load and uncoupled at our shop.

Full sets of data were collected immediately after startup at 0%, 20%, 50%, and 75% load. When the axial movement was observed at 75%, load was decreased back down to 50%, where it remained for the duration of the data collection and continued to warm up with steady state load until testing was completed.

Peak vibration amplitudes on the fan bearings were approximately 0.17 in/s in the horizontal axis on both the DE and NDE. Zip file with full spectral data included in OP (with a couple correct figures in the following post) if you're interested.

-Sean 

Add Reply

Post

HACKATHON in association with IMC-2019

×
×
×
×