Resonance of motor-fan system 4

[ArnoVonck]

Hi everyone,

I'm a student who's doing an internship with IMEC, Belgium. I'm studying a fan powered by a motor. This fan-motor system goes in a resonance at a frequency between 30 and 40 Hz. The operating range of this fan is 30 to 40 Hz so we can't just skip these frequencies. We've tried multiple things to get the system to stop going into resonance but nothing seems to be working. We've already realigned the axis with the motor. (Motor and fan are directly coupled) We also tried replacing the impeller with a new one, replacing the bearings, adding strengthenings to increase the stifness and we even put better vibration dampers under the installation. I dont know what I can do more, so I'm asking you guys for help. Feel free to comment with your findings. Thank you in advance!

Arno Vonck

 

[drawoh]

ArnoVonck,

You can change the resonant frequency making the mount stiffer.

--
JHG

 

[waross]

Have you considered altering the discharge ducting?
If the resonance is flow related, the frequency will be inversely related to the distance to the first elbow or restriction.
Are you able to determine if the origin is the motor or the fan?
Forbidden frequencies or speeds are common in variable speed centrifugal pumps.
Good quality VFDs often have a provision to program in forbidden frequencies.
With a changing input signal, the VFD will not output a steady forbidden frequency, but will accelerate or decelerate through the forbidden frequency.
Another possible solution to flow related resonance may be to change the number of impellor blades.
A non-symetrecal 5 blade impellor may help. That is a two blade impellor mated with a three blade impellor.
In a fluid pump, when the time for a pressure wave to reflect back from an elbow or restriction matches the ime period or multiple of passing impellor blades, you have pressure piling and reinforcement.
It may happen with a fan. I don't know.
If this is happening, it may not be cured by the cures for mechanical resonance.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[ArnoVonck]

Thank you for your answer Drawoh. I know about how stiffening the mount will change the resonant frequency. We already tried this though. We put multiple very strong bars of steel between motor pedestal and the snail house (this is what we call the housing in which the impeller rotates. To stiffen the mount even further we'de have to change the entire unit.

Thank you for your answer waross. I have not considered altering the discharge ducting. The media running through this fan and ducts is NaOH which is causti. So it might not be that easy to alter the ducting, I will pitch this idea to my supervisor though. We have very good VFD's, but the forbidden frequencies at which this unit goes into resonance are the frequencies in which its operating range lies. We need to be able to control the volume flow rate with these VFD's and so that is not possible with these resonant frequencies. I'll look into the 5 blade impeller though. Would it be handy for you guys to have an idea about the situation if I uploaded a picture of the unit here?

Thanks guys!

 

[ArnoVonck]

Sorry Waross,
I forgot to answer your question about whether the origin is the motor or the fan. We've tested the fan to see if he maybe goes into resonance and takes the system with him. This wasn't the occasion. The motor and snailhouse seem to bend into eachother after looking at it with a motion amplified camera. The motor bearings and shaft already have been replaced but it didn't do the trick. I don't think the problem lies with the motor or fan itself.

 

[Gr8blu]

ArnoVonck Usually, a motor is designed to minimize specific (known) mechanical frequencies as part of its construction. For example, the resonant frequency of the frame structure should be somewhere above 15 Hz, since almost all mechanical processes have some form of resonance in the 9-13 Hz range. A reputable manufacturer will also design such that resonances from line frequency (50-60 Hz range) are minimized. However, a VFD puts out a LOT more harmonic content than a "standard" line, so the likelihood of hitting a resonant condition in the mechanical train goes WAY up - particularly if the motor is not originally designed for VFD operation.

Dumb question #1: In which direction is your resonance most prevalent - parallel or perpendicular to the shaft?
Dumb question #2: What is the actual operating speed (not frequency) of the machine? A change in pole count might get you far enough away from the resonant condition by requiring the drive to operate at a different output frequency.

Converting energy to motion for more than half a century

 

[ArnoVonck]

Hi Gr8blu,

The resonance is most prevalent parallel to the shaft. The nominal operating speed of the motor is 1477 rpm, the nominal speed of the impeller is 1476 rpm. The resonance starts to occur in the area between 800 and 900 rpm though.

 

[Gr8blu]

ArnoVonck One more dumb question. If the system is in resonance and the electrical power is turned off (not just reduced to "zero speed" setpoint, but actually "OFF") does the resonance go away or at least see a significant reduction in amplitude?

If the resonance stays approximately the same with the power "off", it is symptomatic of the mechanical system (fan, impeller, discharge ducting, etc.). If it goes away with the removal of electrical power, it may be related to alignment in that the fan/motor combination is experiencing some axial thrust from the impeller that the motor magnetic field is not strong enough (at the reduced frequency condition) to withstand - thereby leading to a "hunting" resonance.

By the way - 800 to 900 rpm on a nominal 1500 rpm machine means you're seeing the resonance between 26.7 and 30 Hz input frequency (800/1500*50= 26.7 Hz. 900/1500*50= 30 Hz). Not somewhere between 30 and 40 Hz.


Converting energy to motion for more than half a century

 

[ArnoVonck]

Thanks again for your answer Gr8blue.

Yes it's been tested, we've concluded that it's a mechanical issue.
Yes, my mistake! Actually the system goes into resonance at 1088 rpm -> 36,2 Hz (VFD frequency) which corresponds with a 18,1 Hz impeller frequency.

 

[waross]

Simple solution:
Put a baffle in the fan discharge and run the system faster.
I once encountered a fan system that was overloading. (Design error)
The millwrights cut a slot in the discharge duct and lowered a plate into the slot to restrict the discharge.
As the flow was restricted, I monitored the current.
When the current reduced to rated current, the millwrights welded the plate in place.
It wasn't pretty, but it was quick, cheap and effective.
You can use a baffle to reduce the flow, and then increase the speed above the problem speed.
Or, you can trim the ends of the impeller blades shorter.
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[ArnoVonck]

Thank you waross! This seems like a good idea. I'll run it past my supervisor and keep you up to date.

 

[dvd]

It may be irrelevant, but here is an interesting bit of information on Helmholtz resonance for pressure blower ducting -