Dynamic Balancing at low speed. Will it work at running speed ? 5

[edison123]

We do dynamic balancing of motor armatures, steam turbines (rigid rotors not flexible rotors) etc. in a hard bearing type dynamic balancing machine at 300 / 600 RPM depending on the size. We follow ISO G 2.5 std. and go below the specified limits.

Will dynamic balancing at a low speed will ensure the dynamcially balanced state at the runing speeds which are typically 1000 to 3000 RPM in motors and 5000 to 10000 RPM in steam turbines ?

 

[anands78]

Yes for rigid rotors and No for flex.
When we balance a rotor we choose a speed lower than the critical so that we get an accurate phase(at speeds close to the critical the phsae changes rapidly with speed). By doing so we achieve the objective of making the rotor pass the critical without a large amplitude of vibration.

In a flex rotor, this is not the case. The mode shape is different for the first critical(half sine) and the second critical(full sine). In this case the balancing for the first critical will actually cause a huge imbalance close to the second critical due to the shape of the rotor. When bent as a half sine, the bent part is above the centre line, but in a full sine half the shaft is above and the other half below.

 

[gunnarhole]

Edison123,

The motors will be OK for standard designs of under under 2 MW. At least I have never seen a motor smaller than that with a bending critical under 50 Hz. I suppose a longer rotor, like you might find in an explosion proof or submersible motor might but I don't think so.

As for the steam turbines I think that there are lots of them with critical speeds in the range you mention. Whether they need to be high speed balanced is questionable. I doubt it, but I would put it to the OEM.

The only rotor I have high speed balanced was a large axial compressor. That was done in a vacuum chamber and took over 24 hours to complete. This was due to the cycle time for evacuating the chamber and an unfortunate combination of available balancing planes. Luckily, this is not often necessary.

Regards,

Gunnar

 

[kimbo1]

A couple of good answres above. Simply though, with a low balancing speed, the accuracy of the balance machines transducers becomes more critical as the out of balance force you are trying to measure is very small. As long as the machine is properly calibrated it will be fine.

 

[Ralph2]

To balance on a hard bearing balance machine all that is required is enough speed to create a measurable and consistant(stable) force.
Check it out yourself. Balance your rotor at 300, then take it out, reverse it and check it at 600. If your machine is properly calibrated there should be very little difference in the resulting unbalance <5%. If there is more you have a flexible rotor or something is wrong.
Good luck
Ralph

 

[Ralph2]

Further to my response I should have qualified it with. You must be running at a speed relative to the machine sensitivity. On our Schenck HB-5 for example if my tolerance is >20 gram inches I can run from 120 to 359 RPM; for 2 gram inches 390 to 819; .2 gram inches 820 to 1799 and at .02 gram inches 1800 to 5000.
This has never been a problem in my experience because I seldom have tolerances in the .02 range with the size of our machine. Fitting even a 2 pound rotor into a H5 balance machine is problematic where even then I would have a tolerance of .03 gram inches per plane (API 4W/N at 3600 RPM)
Ralph

 

[electricpete]

How about a 10-stage vertical centrifugal pump with metal bushing type bearings between stages. In considering whether that is flexible, should we only consider the distance between bearings?

 

[Ralph2]

I would say yes. That is where the physical restraints to the shaft are. In theory the rotor should never touch the internal bushings. And, when it does, it is cushioned to some degree by the liquid being pumped.
In my experience this type of pump is difficult to balance because of the repeatability problem. Take it apart, do nothing and re assemble and you will likely find a large discrepancy in the unbalance from the initial run. Caused 9 times out of 10 by non square faces of the impellers and spacers. The tighter one does retaining nuts the more the shaft will bend thus changing the centerline of the rotating mass.
For what it is worth, in our shop we would balance each impeller separately except for the two end ones. Then assemble the complete rotor. Check the runout, if it changes depending on how tight the retaining nuts are then back to the machine shop to face each side of the impellers and sleeves (mounted on a mandrel so that both sides can be machined from the same setup.
Only when tightening makes no difference will we then balance the element. Removing material from the end impellers to remove the &quot;couple&quot; and spaced along the middle to remove any static.
And, check it at several operating speeds to ensure that the rotor is not significantly flexing at speed. If it does then it must finally be balanced at the operating speed.
Good luck
Ralph

 

[GregLocock]

That's good practice and is something we've only started to investigate recently - what variation do you get when you reassemble the same system from the same components? If you haven't got a handle on that how can you justify rebuilding with different components to test them?

I have generally found that with shortish simple systems that the balancing speed is immaterial, but with more complex assemblies such as complete vehicle drivelines it is vital that you balance at, or near, the problem speed. Balancing at/near resonance introduces its own special problems, but generally the advantages outweigh these. Cheers

Greg Locock