Duct Mtd` Fan Motor

[stardelta]

We have presently in our shop a Woods Fan motor, its pad mounted (8 rods) in a circular duct, fan size is approx 1225 mm dia. Its a 3 ph 2 speed single winding 6 and 12 pole, max Kw is 7.5. The customer is complaining of abnormal noise/ vibration when running. we have been to town on this, all housing and journal sizes checked, new bearings, fan and rotor balanced as a pair, air gaps/concentricity checked and the winding has been surge and load tested. We still cant find a thing wrong with it. There is a lot of electrical noise when running on the 12 pole winding but thats to be expected. GSE is 0.020 on each bearing and we cant find a vib` figure higher than 2.2 mm/s anywhere on the motor or outer enclosure including the duct yet it sounds like a bag of nails with and without the fan fitted. Our only conclusion is that its vibrating at the natural frequency of the duct and the motor mountings. The customer says its always been noisy but in the last 4 weeks its got steadily worse. Can anyone offer any advice on probable causes?

 

[electricpete]

Tough problem. I don't work with audible noise much and don't have much idea.

My only thought is that audible noise is often associated with stuff showing up at the higher frequency vibrations, perhaps slot harmonic pattern. If you've checked concentricity that's about all that can be done to address electrical noise. Some motors inherently have a lot of electrical noise. I have heard it can be associated with fracitonal slot design. Also since it is flux-dependnt perhaps it might be aggravated by high or unbalanced voltage (longshot).

Might also poke around in the spectrum to see what is up there in the 20kcpm -200kcpm range. Maybe a fan-blade pass frequency? I don't know if you have capability to record and get a spectrum of audio data but it seems that would be helpful in narrowing the source.

Out of curiosity, why do you say that high electrical noise is expected on the low speed.

 

[stardelta]

Pete
Thanks very much for your comments. I spoke today with an in-situ balancing `expert` who seems to think the problem is related to resonance at the natural frequency of the motor and duct combined, and what he terms the air-pulse effect of the fan. He is pretty confident he can solve the problem by careful trimming and by adding mass-dampers if required. I cant comment as its not my really my field but it has been backed up by our condition-monitoring guys. I will let you know the final outcome. By the way, my term Electrical noise describes the `Hum` for want of a better term, and vibration generated by slow speed AC windings when energised. Basically the more poles the winding has, the more they have a tendancy to make a noise and shake!

 

[GusD]

Hello Stardelta.

Sounds like a tough one.The Vib levels you mention are not abnormal by themselves ,yet; unit (fan and Duct work) still vibrate excessively.Costumer says that it got worse since 4 weeks ago.I suspect something had to happen to the"ductwork".Its natural frequency may have shifted due to stiffening/added mass/looseness or any other work that may affect its behaviour.I would make sure that duct work is as it was a Month ago.If it is,I would perform "bump tests"on the structure,and see if motor/ fan vibrations are exciting it in some way.
I can't really think of anything else that I would do in your case.
Stardelta,you mention Hi vibraions,but you don't give the predominant Frequencies causing your problem.They might show some connection.

Good Luck

GusD

 

[electricpete]

how can it be an air-pulse effect if it vibrated "like a bag of nails" without the fan. I assume you were running the motor mounted in the duct without it's fan blade at the time.

 

[SteveKW]

The fan hub assembly, fit and/or a blade can make a nasty noise if they are loose!

 

[jclough]

I've had good luck fixing a similar problem. I altered the motor speed slightly by honing and adding shims to the bearing bores. Changing the bearing preload altered the motor speed 5% and quieted everything down.

 

[electricpete]

how does adding shims to the bearings bored alter the motor speed?

 

[GusD]

Hi stardelta

To jclough.I don't know how big a motor you did all of that work on.But that was a very corageous move.
Unheard of possibly.If what you claim to have done,didn't happen to work out,what then ? Throw the motor in the garbage.I believe you were extremely luck to somehow solve a problem by a very unorthodox and risky move.It does not happen like that most of the time.
I would not recommend that solution to solve any vibration or any other problem for that matter.
It worked for you and that's great.I doubt it if you could get lucky twice.



GusD

 

[stardelta]

Hello guys and thanks for the comments. We have at last cured the problem to the satisfaction of the customer. We decided to carry out a NDT test on the metalwork of the duct, as it was covered in several layers of Hammerite paint it was shotblasted. This revealed weakneses in the areas of the rod mounting holes, in 2 positions the metal was paper thin. We bit the bullet and obtained a new duct and after fitting the motor and fan the diffence is amazing! Its certainly something we will be checking in future when dealing with this type of machine.
A question for Jclough, how big exactly was this motor you are talking about? Like Gus I cannot believe this method can be succesful without compromising the relability of the machine.

 

[kamalpanda]

dear stardelta,
its nice that you fixed the problem.So the noise was due to the mechanical resonance in some respect.

I read your thread today.before you mentioned the noise is due to electrical reason.I hope in such a doubtfullcase its better to perform the solo running of the motor in the work shop without the duct(if possible)

anyway the problem is closed now and thanks for letting the forum know about the remedial actions taken by you.

 

[electricpete]

I have a question on whether we conclude this was a resonance problem or looseness problem.

If resonance, I have a hard time understanding how it doesn't show up on overall vibs but makes a significant noise. I'm guesssing a structural resonant frequency is fairly low <3000cpm and human ears simply aren't very sensitive to low frequencies (but accelerometers do great).

If looseness, I would have thought there might be some indication from vib (high demod energy, impacting in time waveform), although maybe attenutated because the machine and it's mountings are buried within the duct.

 

[electricpete]

By the way, thanks for providing the feedback.

 

[SteveKW]

On one particular brand of motor I repair, I often have to increase the preload of the bearings because of the noise! Although I doubt that it lowers the RPM of the motor as stated by Jclough. Typically, I just increase the bends in the wavy washer.

 

[jclough]

Gus and electric pete,

The fans on which this method works on are small 400 hz 2 to 3 inch diameter blowers which are used in certain aircraft. Part of the test spec is that the motors must achieve a certain speed within a specified time-frame. We have to run the motors for a period to get the grease warmed up before test. lube amount and viscosity are important here.

I've also had excellent result on some of the brushless and slotless brushless motors we make.

Pete--

How does it work?

We use LabView to calculate prelooad as true bearing preload is at .000&quot; deflection. Altering bearing pre load changes the frictional losses in the bearings. There are however limits of course. Too little and the bearings start to make noise. Too much and bearing life (as well as speed) will suffer.

Cylindricity of the two bearing bores is also important in this application. In order for the preload measurements to be accurate, the rotor and bearings must be free to move in the endcaps. We all know what kind of problems misalignment can cause.

Gus,
Necessity is the mother of invention.
One of the most valuable things about this board is the wide variety of backgrounds and experiences.

Often when we change parts we assume things like bearings are the same size or the bearing bores are round.

Endcaps for instance usually have a three point out of round condition due to the chuck pressure vs. tool pressure on the machine that they were made. They look fine with a plug gage or inside mic. but forces to to loading or unbalance quickly pound the lobes until the endcap is oversize.

If it doesn't work? ... Usually I'll try altering the rotor skew a little. You don't even want to know how I do that
Sometimes taking another .0001 or so out of the stator bore helps.More airgap--> more speed to a point.I may also try altering the shims in the endcap opposite the spring. Axial position of the rotor can also affect speed.

I'll discuss further if anyone is interested.

But you guys probably already knew that.

Regarding the use of such methods, unorthodox? maybe. risky? probably not... it was junk when I started.

Jon

 

[electricpete]

JC - Thanks for responding. I don't doubt what you say but just trying to understand. How do the shims change speed. Are we changing the friction on induction motor enough to significantly alter the slip?

 

[GusD]

Hello Stardelta

to Jclough.Jon ,I am with you on &quot;necessity being the mother of invention&quot; .As you say,this board is great because we all have a chance to learn from each other and benefit tremendously from it.These Forums are the best University,that anyone could attend without leaving the house.My questioning of your options of previous post,had to do with the specific piece of equipment we were dealing with, rather than the method used to solve theproblem.Many years ago,I was rewinding Dynomotors 400hz for the air force.The little armatures sometimes had 2 or 3 sets of windings and the balance specs on those things were so tight that you wouldn't believe.I did come up with a few tricks of my own, to stop throwing 3 away out of five.
You probably wouldn't believe what I had to do.Just like yours,this was a specific piece of equipment that operated under different set of rules.Never let go of the inventive instint that solves a problem.It is great that Stardelta got his problem solved and we all benefitted from the interaction.Jon,Stardelta,Epete,Jbartos and all the others,please keep it coming.

Thanks

GusD

 

[plasgears]

I would favor the stiffening of the mount on the duct, then isolating the fan mounting from the stiffened duct interface using rubber pads. In addition, duct panels near the fan may be excited, so some dampening is in order. I have used adhesive mounted viscoelastic panels, which have a great deal of internal damping.

 

[jbartos]

Suggestion: The motor noise may have electrical origin. The power supply quality might be checked for harmonic content. The noise of electrical origin may impact mechanical hardware or deteriorate the mechanical support too.