diferent vibration values over time - further data 1

[Esbirinai]

I checked a ventilator (100 kw - 960rpm) and obtained different values of vibration. Values go from 4 mm/s to 11 mm/s. The ventilator has 8 blades (2*4 blades, two on 0º, two on 90º, two on 180º and two more on 270º)
The spectrum returns only a peak near 64Hz which i presume to be from the blades (960 / 60 * 4 = 64 Hz) Actually, this peak goes from 4 to 11 mm/s in separate measures.
I do not understand. Nothing changed between the measures.
I can guarantee that there is nothing wrong with the data collector and probes.
Actually, There were two fans on site, they are exactly the same (axial flow fan). One of them is ok (rms 3 to 5mm/s) and the other has symptoms like the one i described.
The inlet flow is not varying on any of them.
However, with a close inspection i found that the blades on this fan were slightly bended on one end (the blades on the other fan were not bended) and that a small part of one blade was missing.
Could this be the reason? The bended ends are enough to make this happen?
Thank you for any help.

 

[hacksaw]

Evolving data is typical of progressive failure.

You need to secure the fan and give it a thorough inspection.

 

[vanstoja]

Your description comes across to me as a two-row impeller with 4 blades per row that are staggered by 90 degrees between the two rows. This is a concept that has been tested and has been used infrequently in centrifugal fans and pumps to minimize impeller blade passing (IBP) vibrations particularly when the downstream volute tongue or diffuser pressure recovery vane inlet edges must be close to the impeller blade trailing edges. I never heard of this concept being employed for axial-flow fans or pumps. Staggering the blade trailing edges in two adjacent rows has been claimed to produce an acoustic cancellation effect downstream of blade trailing edges and to raise the fundamental IBP frequency to double the number of blades in each row times rotating speed. Neither claim was substantiated in tests of a centrifugal pump impeller by one pump manufacture whose data was available to me. I don't understand how or why such a concept would be used on an axial flow fan which wouldn't or shouldn't need a cutwater downstream of the last row of blades that would cause blade-vane interaction IBP vibrations. If, in fact, you have a two-row, staggered blade centrifugal fan, then bent trailing edges and missing parts of one blade could indeed elevate IBP vibration levels either by hydrodynamic interaction with the downstream cutwater or by unbalancing the driver's rotor assembly and increasing the fan's orbital radius thereby increasing the cutwater passing fluid pressure pulsation intensity .