Overhung Fan High 1x Axial at 90% but then drops slightly at 100% (resonance?)

Look at your spectrum and all those harmonics, Can you get a waterfall plot of the axial fan nde?

@fburgos posted:

Look at your spectrum and all those harmonics, Can you get a waterfall plot of the axial fan nde?

Hiya, unfortunately i don't have the software to be able to create a waterfall plot. i think it would look rather like a standard resonance graph where there would be a dramatic high amplitude in the axial axis versus the others. i guess changing the Fans damping would alter its natural frequencies anyway right? 

AFAIK, resonance dont have harmonics.

This problem is present since installation, then presumably its a new machine, yes its posible a resonance, also posible its a installation problem, a damage bearing during transportation, what does the OEM said?, do you have spectrums to compare from new to now?

you dont have axial from DE bearing, can you get that one, do a phase analisis and post results here.

Some posible ways of investigating resonace is to do a bumptest to the the machine, get phase data of 1x during castdowns, measure 1x across the bases looking for discrepancies across baseplate.

sorry to say this, but from overalls and 1 spectrum no way is posible to give an advice.

Dear James, 

Resonance happen when one frequency of machine (in this case 1xrpm)  is almost same that one natural frequency in the base plate, or one component near to the fan. bump test is recommended in machine points , in the same direction of highest amplitude. 

if the problem is resonance, the amplitud will change when you modify the rpm.. I  believe is not our case.

the flow can create variation in amplitud if you have looseness, please, if you have two channel data colector make cross phase analysis, near to he high vibration point. 

next step is check internal cleareances, run out of shaft, coumpling condicion, an and try to make balance of rotor.. 

if turbulence is affecting the system, check looseness in damper (air inlet flow control system)

some cases are not easy to find de cause, but if there is one problem, also there is one solution. 

Regards. 

Hiya

Thanks for the reply. Firstly I thought it was resonance because of the high 1x which does change with the speed of the fan, however I don't know enough about the role of air flow on vibrations and if that would make a difference if for example the air flow inlet was restricted or if turbulence was an issue..

I'll do a bump test and try and check internal cleareances, run out of shaft.. 

Cheers 

Hamish 

Hi guys I thought I'd add a video of the NDE bearing and the amount of movement in the axial direction it created at 90%.

Greetings JamesPeters ,

i dont think its a flow related problem ,Better you have to check belt pulley sheaves V-groove profile ,alignment ,eccentricity & belts tension  .

Hi Kishore Kumar

Oh right that's interesting! Do you think that the belts and pulley arrangement is able to create a high 1x peak at 90% but then reduce down at 100% running speed ? I would have thought it would just get worse is pullets weren't aligned or the pulley was eccentric

Is that the "fixed bearing ?

Do you have phase measuring capabilities  with you analyzer?

In the video I //think// I see the pillow block wobbling axially.

I'd be very interested in the contruction of the fan pedestal top plate.  Details like  thickness, and confirmed presence of any welded stiffeners in close proximity to the bearing bolts to start.

I'd also take a bunch of spectrums and 1X phase vertically all over the top plate.

Hiya Dan

So this is the floating bearing , the NDE bearing. And unfortunately no I can't get phase readings from my vibration sensor which is rather annoying.

Yes you're right, from the video it looks like the pillow block is moving axially back and forth with the shaft. 

Okay I'll see what I can find out as I've not got that information to hand. 

I would agree that an resonant structure in the fan axial direction with frequency near 90% of full speed is possible. The "deck of cards" shim pack under bearing housing and possible thin support base are suspects for correction.

The video indicates a rocking vibration mode. The low visual frequency is related to camera/phone video rate; correct?

Walt

Hi Walt 

Thanks for the advice, I've added a couple more videos for people to have a look at. They are slowed down so it's more visible to see the amount of axial movement going on.

As stated by others, resonance is typically speed related, not present across the entire speed range

The plates on which the fan bearings are mounted is typically twice as thick as the actual top plate. The top plate is not thick enough to support the axial movement. Both videos you provided clearly show this axial movement

By the looks of the fan base, there is very little stiffness on the top plate. I have had success by putting "C" channel the entire length of top plate, from each bearing bolt on the underside of top plate, parallel with the fan shaft. The "C" channel must fit tight to underside of top plate.

IF the fan is new, I would be talking with the fan OEM to get them involved in their design flaw.

Dave

"As stated by others, resonance is typically speed related, not present across the entire speed range"

Hi Dave glad to hear from you, okay so from what you are saying I don't understand why the initial vibration peak changed after additional duct work was added. It initially had a vibration problem that was worse at 80% and then after the additional ductwork was added it changed the point of the worse vibration to 90% so to me that would have suggested a speed related resonance issue but due to the change in mass it affected where the natural frequency was?  

Yes you are right  the videos show alot of axial movement, I didn't realise that the thickness of the top plate would have that much of an effect but now that you mention it it does look rather thin. Would this likely be the main cause of the axial movement, a lack of stability in the top plate ? Yes the fan is relatively new ( one year old)   the manufacturer is involved so I will suggest someone having a look at that as an option.

Thanks alot 

James

In my past, I had a large overhung fan such as this where the base was insufficient and allowed the fan to oscillate in a sine wave pattern.  bearing life was horrific.  Torched a hole near the top of the hollow base and pumped it full of concrete to stabilize it and the issue disappeared.

The last video shows the DE bearing housing is moving along with the support base, and it is not rocking much. While resonance is probably present, the following produce the dynamic forces:

Fan sheave axial runout (wobble)

Motor-fan sheave misalignment

Fan rotor axial runout (wobble)

Fan rotor dynamic unbalance (2-plane)

Walt

Check for Soft Foot and a twisted frame caused by heavy ducting. Also belt alignment 

Adding mass lowers natural frequency, adding stiffness raises natural frequency, sounds like new duct added more stiffness than mass

Other fan mfg companies have this same issue with thin fan bearing mounting plate, if you have the ability to collect phase data, you should see the plate is moving 180 degrees if truly moving

Dave

your videos are nice to show its moving, but at what frequency? please post spectrums and phase data.

If the fan is supported on isolators, check if the selection of the isolators is correct (stiffness, qty, position) paying attention to the center of gravity.

Flow-induced vibration could be the main problem, can you get some graphs of air temperature, flow, rotating speed, and vibration overall level on one graph?

Hi FBurgos,

I've attached some spectrum data for you to have a look at if you want to. unfortunately i don't have the facilities to collect phase data so can't provide that at the moment. 

thanks for everyones responses. 

i will check the vibration mounts and see if we have any further details regarding air flow and temp

there you have two spectrums showing harmonics up to 20x, i would fix this loseness first.

Please tell us what the fan is doing? Any chances of deposition? What manufacturing plant is it?

This is definitely a structural issue.

I faced similar issues with fans in paint shop of a car manufacturing indsutry.

because of the poor base any time there was paint deposition on the impeller the vibration shooted like anything.

We kept on doing impeller cleaning & balancing, did result in the vibration reduction, but after 10-15 days again the same problem.

In your case its the looseness getting excitation because of structural weakness.

If you correct the looseness the vibration might reduce but next time any other issue will get amplified because of the poor base.

Do correct the base. Strengthen it rock solid.

As discussed above the possibility of resonance in the axial direction, run/up coast down is crucial, also if the motor is powered by a VFD, you can vary the speed (higher/lower) from the current speed and check if the axial vibration is dropped.

In many cases, high vibration due to unbalance causes transient events in the time waveform that translates into looseness type 3, it doesn't mean you have a looseness problem. When levels are reduced after a balancing correction (correction away from the resonant frequency), harmonics will disappear.