While trying to balance my 1x was changing from 0-6 mils (phase shifted slightly). How can you balance with this? I worked through it and balanced and it seemed ok for a day then levels jumped up again. It would seem that something else is wrong in fan. Overall velocity beating also. Fan cracks??Structure cracks? Looseness? Bearings seem ok, good levels..

Any ideas

Do you have another "fan" nearby that runs about same speed? You may have beating going on
Your statement that phase shifted "only slightly" - less than 20 degrees??
If so then this may cancel out possible rub.
Are the fan blades clean??
There are many other possibilities....

quote:

Originally posted by jbalko:
While trying to balance my 1x was changing from 0-6 mils (phase shifted slightly).

If so, the culprit could be IMO structural instability, such as cracks, looseness. Even if there was a nearby equipment running at close frequency it should not affect phase as it is referenced to this machine speed. This is assuming phase was near steady.

If phase was fluctuating much wilder, of course, resonance at 1x could be a suspect.

Question:

Was there beating which produces periodic amplitude variation or just instability having random amplitude variation?

I suggest setting up a long time duration test to sort out between those two. I have had a case where structural cracks produced random amplitude variation in a wide range with slight (5-10 deg) variation in phase.

David

The day after balancing, I was called and told the fan was vibrating bad again. Except when I got there it was ok or the same as when I left it. I balanced levels even lower. I tached in fan speed and belt speed. Fan speed 777 rpm. Belt speed is solid 388 rpm. This may account for the shifting (2x this with fan speed). But I found the major vibration at 725 CPM. I have no idea where this is coming from. Even though the mils level is down below .5, overall velocity fluctuates between .17 and .27 in/s. Mainly at the 725 cpm freq. Bump tested all over the unit but cant identify the 725. Now we just hope it stays running ok. BTW Never did this in the past, just started.

This message has been edited. Last edited by: jbalko,

725 cpm - that is weird. Just to fill out the picture, what is motor speed? Does that frequency show up on the motor? Does it show up on the base (perhaps coming from other equipment)? Might be interesting to look at the belts/sheaves with a strobe for anything weird.

quote:

Originally posted by jbalko:
Even though the mils level is down below .5, overall velocity fluctuates between .17 and .27 in/s. Mainly at the 725 cpm freq.

Doing a quick conversion, 0.5 mils pk/pk at 725cpm would be around 0.02 ips pk/0. Was the mils measured on the shaft or the housing?

Yes its in the motor. There are harmonics of it in the motor and fan. Did not check base.

Brand new belts and sheave. levels fluctuated before and after new belts and sheaves.

Motor speed 1186.

.33 mils at 777,, 4-5 mils at 724.

Most machinery shut down while balancing. No other units directly near by, though piping could carry freqs. When fan off freq not there.

what type of fan wheel is it? what environment is it operating in? I had an airfoil fan that is operating at our recovery boiler acting simularly. Phase was constantly changing while monitoring. what was found was that condensation had formed inside the blades. Drilled weep holes in the trailing edge of the blades and released the water. Phase stabilized and I was able to then balance the fan.

Scott

If phase is changing,continously on a specific angle in a cyclic manner(example:60-100,100-60,60-100,100-60).I trust i have not confused you.

Then you can't Balance the Impeller, this may be due to some looseness, may be due to Keyway or some sort of looseness. First correct this problem and reasses the unabalance.

Good luck.Expecting your findings.

jbalko,

So, it appears that the problem is not at the run speed of 777 RPM, but at a mysterious 724 CPM. Is 724 CPM peak consistently present or fluctuating in magnitude? Is there enough resolution to make sure that no two frequencies are contained in this bin? Could you please post some data?

David

Here are some plots. When balancing, if averaging was used, phase and amplitude stayed relatively stable.

fanplots.doc (128 Kb, 35 downloads)

This starts getting interesting!
It appears from the spectrum and TWF that 724 CPM is being amplitude modulated at 54 CPM. Is it possible that 724 CPM is actually fan's running speed? Did you check if fan speed stays rock solid?

Are there harmonics of 724 CPM? Can you collect data with better resolution? What is current spectral resolution?

David

While balancing rpm stayed very steady. Kept checking belt speed and that stayed steady.

There are more harmonics of this freq, than running speed. Again I will repeat, I check this speed many times, for periods of 30sec to a minute, with no change. Motor shows harmoncs well too. Motor is mounted on side of housing.

Current spectrum is 0-1800 / 400 lines

I will not be able to recheck this again unless it gets real high again.

plots2.doc (82 Kb, 19 downloads)

Belt resonance????

jbalko,

Isn't 724 CPM by any chance belt frequency?

The belt speed is a solid 388rpm, tached in with tape. Also sheave ratio and centerline confirm this. 2x belt is 776. I corrected the earlier belt speed. When I used the strobe the 1X fan and 2X belt were .5 rpm apart.

Is there dampers done the line that open and close? Is it are turbulance?

Could it be a bearing frequency modulated with the cage frequency.

Hi,

I am having a more or less similar balancing problem on a 26 MW generator assembly. i balanced it once... shaft and casing vib levels came down. after 15 days vib levels went up. but there was once a sudden load reduction/throw from 14 MW to 0. after which the shaft and casing vib levels went up. now again i am here for rebalancing.

Gen DE and NDE X probe values are consistent. Gen DE and NDE Y-probe values are hunting by 20-25 microns. what may be the reason before i post the signatures...??? i will post the details and spectrums in a day or two...

Madhu