Higher Vibration Problem

My first guess is there's a structural resonance at running speed in the fan bearing support structure...amplifying vibration in the horizontal vibration. Confirm by examining the force response function plot from an impact test.
Second guess would be pedestal looseness. Confirm by running fingers along bolted or grouted interfaces in the pedestal...checking for looseness or "nibbling".
Dear Max ,
Thanks , I also feel it as unbalance problem . but why the vibration at fan free end ie location no. 4 is reducing once the rpm is increased beyond one particular rpm as shown in the data sheet . I think the vibration due to unbalance should increase with increase in rpm. In this case , it is increasing at fan drive end ie location no. 3 with increase in speed . But at location no. 4 , the vibration is increasing up to one particular rpm and then reducing to satisfactory value . I am not understanding this phenomenon at fan free end bearing ie location no. 4
If balancing is to be done , at what rpm I should try balancing and in which plane - drive side or free end side .

No other data is available at the moment .

I tend to agree with Georges comments earlier that there is another fault which is amplifying any unbalance existing in the fan. Like any balance job you need to eliminate any other possible sources of vibration, such as alignment, mechanical looseness,resonance ect. I would check for all these faults and then try to correct any unbalance at the problem speeds on both planes if required.
After I did a bump test to help confirm these readings and made sure nothing, like hold down pedestal bolts were loose, I would do a 2 plane balance at the top speed of the fan and see what happens.
You haven't mentioned phase relations between the two bearings, are they in or out?
I see that the axial on all the bearings is fairly stable through all speeds, but #4 bearing is slightly higher than the rest. Have the dampers on each side been check for correct "pull" or does the air flow get regulated by the speed of the fan?

Does it have a history of this problem going and coming or is this something new?
I agree with Ralph. Instead of trying to guess with insufficient data, you need to do a phase analysis. Then you will know for sure. Unless you have a history of knowing problems that develop on a specific machine and the vibration characteristics that accompany those problems, jumping directly into doing a balance correction is usually premature. If you are unfamiliar with how to interpret the phase data please post it and we can help you with that task.
Dear Ralph , Thanks , Regarding 2 Plane Balancing , the diameter to length ratio of the fan is large . Please tell me why we should go for two plane balancing in this case .
Regarding phase measurements , right now i am unable to measure bcos i need to stop the fan to apply some reflexive media on the shaft . I am waiting for the opportunity to stop the fan.

Regarding history of the fan , The impeller is newly installed . Fan casing is old .

Regarding air flow control , mainly controlled by speed . Dampers are also provided at both sides and they are 100% open .

Please confirm how did u perform the bump test with these readings .

Regards ,

the diameter to length ratio of the fan is large . Please tell me why we should go for two plane balancing in this case .


In theory the wider the fan the more probability there is for 2 plane balancing, but I have seen center-hung (ie: 8 feet diameter and 8 feet wide) fans which had close phase relations between the inboard and outboard bearings and a simple single plane balance would work if the weights were placed in both planes in a ratioed position and size. Cuts down on total number of runs and time.

You call this a "Preheater Fan", does this mean it blows air into a preheating system or moves air that has already been preheated?

You say this is a new rotor, has it had this problem from the initial startup?

Please confirm how did u perform the bump test with these readings.

You do not perform the bump test with these readings. Using the bump test, or coast down plot, you check and see if the 2 bearings each actually do have a different natural or critical frequency at or near the same CPM as your posted readings. If not, then something else is probably causing the phenomenon.

Where is the held bearing located, coupling end or "free end". Are we sure both are not held? Is the floating bearing in such a position that it can indeed float? Do you call the "free end" free because it is where the floating bearing is?

I may have missed it and you may have already said, but what type bearings are they running, antifriction or sleeve babbitt type?
WHILE a rotor pass throught a resonance zone the phase change is 180 degree. If rotor is in resonance, when you install phase reader, data taken has tobe unstable.
About increasing and decreasing of 1X amplitude in bearings 3 and 4 respectively, above 868 RPM I think the following :
1.- LOw damping bearing (i.e. high stiffness) has in resonance fast change of phase angle (180 degree of change) in low change of RPM (bearing 4). The amplitude decrease to low level when increase RPM.
2.- High damping bearing (i,e, low stiffness) has in resonance slow change of phase angle (180 degree of change) in a large change of RPM (bearing 3). The amplitude decrease to medium level when increase RPM.
Notes :
-Rolling element bearing has high stiffness and low damping, and
-Sleeve bearings has low stiffness and high damping.
Question :
Does your fan has rolling element bearing in bearing 4 and sleeve bearing in bearing 3 ?
Conclusion :
In adition of check balancing and shut down record to check resonance I suggest chech radial clearance due to wear.
If you have acces to a 2 (or more) channel spectrum analyzer that allows you to acquire waveforms simultaneously on both channels, you may not have to shut down the machine and apply reflective tape to do a phase analysis. You keep 1 accelerometer in a fixed location and observe the relative waveform position of wherever the other accelerometer is located compared to the fixed sensor. The key is that the waveforms must be acquired simultaneously and the task is easier when the waveforms are close to sinusoidal. I still think a good phase analysis is the best first step in your situation.
Good day guys!
Did you perform the runup /Coast down for this machine..From the data it looks like you have couple unbalance.Do perform dual plane balancing..if your phase 180 opposite at H plane for both bearings. looks like you've wobling at axial too, may from this couple unbalance. This can be solved!!
Dear Ralph,

The fan moves air that is already preheated ( Induced Draught Fan ) .
The problem is there from initial start up .
You are right , Free end means Floating Bearing side . In this case , Drive end is Fixed and Nondrive end is Floating .
They are antifriction bearings ( spherical roller bearings , Oil lubricated ).

I have performed single plane balancing but never tried 2 plane balancing .Could u please give me the steps to perform 2 plane balancing .
Regards ,

I have performed single plane balancing but never tried 2 plane balancing .Could u please give me the steps to perform 2 plane balancing .

What do you have, as far as equipment, to balance with?

I would first make sure the phase is out between the #3 and #4 bearings in the horizontal direction. Like I said, if not, then a ratioed weight amount and position could be done through single plane balancing on a center hung fan.

It is strange that the "free" bearing has such a higher amplitude in the axial direction than the held bearing. I would make sure this bearing ("free") can indeed float and not positioned in the wrong place in the housing.

I am sorry, but I am not familiar with that brand of data collector/analyzer/balancer.
Does it have a balancing feature that does 2 plane balancing or just single plane balancing or either?

Maybe some who uses this brand can direct you in the 2 plane balancing. If not, we'll come up with something. Smiler

Can you provide more information about the fan? Have you tried single plane balance on this fan, and it was not successful? If you have not done a 2-plane balance, then I suggest practice on a small machine. I use a bench grinder with machined aluminum disks in place of the abrasive wheels for conducting training. Why take a long time or make a serious mistake on a large process/production fan?


The fan is raw mill system air fan. 

Rpm 900 

High horizontal vibration is observed with 1*x dominant. Cross Phase difference is 95 at DE bearing.  So we suggest to do dynamic balancing and we got success also by single plane method. I just want to know the exact condition or situation when we have to perform two plane.. 


Add Reply

Likes (0)