Vibration in Turbine Engines 2

[MecArt]

I work for a turbine engine overhaul center. It is known that there are variety of reasons that cause vibration in turbine engines. The main reason is unbalance in rotating parts.In balancing procedures sometimes the unbalance value differs in subsequent runs.In general these differences are small so that they dont effect the residual unbalance value. however I experienced a 2 g.in unbalance value change in two subsequent run despite of nothing changed in the rotor and arbor.The unbalance tolerance was 1.5 g.in in either correction plane.The balancing machine is a hard-bearing horizontal dynamic one. The arbor is a well balanced nut-tightened arbor.Does anybody know the reason for this difference?

 

[butelja]

If the rotor is built up of several subassemblies (as I assume it is), then it is possible for internal friction between the parts to change the miniscule bow of the rotor from run to run. Rotors with many shrink fits and/or bolted connections can be succeptable to this type of hysteresis. About the only suggestions I can offer are: to follow the manufacturer's assembly guidelines, check the tolerances on all interference fits, and torque bolted assemblies in the proper sequence and to the proper torque. Good luck.

 

[ArneLindholm]

Dear mec, butelja has indicated the main aspects. Unbalance means moving of the centre of gravity related to the rotation centre. A these discussions are based on the simple formula = Rotormass times excentricity equal balancemass times radius. (I f you liek to know more check the Standard ISO 1940/1). You can use units in [Kgxmicrons]=[gxmm]. A perfectly balanced disk of say 10 Kg that you can move 1/10 mm (=100 microns) in a radial play will change the unbalance at a radius of 100 mm with 10 g. Keeping the radial tolerance good is the major long term problem in jet engine overhaul. Everybody "must" clean and touch and gradually the plays are opening.
I have a very good experience from some things to care for:
1. Spray or add electrolythically and grind clean all radial clearances to minimum play inside the allowed according to manual or ask manufacturer (watch that you follow manual and if applicable FAA paperwork).
2. Measure the actual runout (dial gauge) when you balance the disk to zero unbalance. (If possible both sides, min. one side). (It is always some amount. Record how much and the direction!!).
3. Build the final rotor so that each disk has opposed or unbalances distributed in a spiralized fashion.
4. Check the actual mounted runout disk by disk and compensate for it when adding next disk as much as possible. Keep a careful record of the full rotor.
5. In cases where you are allowed: Make the fit almost zero and heat the rotor part to
somce 50K temp above room temp before the mounting.
6. When finally trim balancing the full rotor, always remove the static unbalance along the centre (spread any grinding well, smothly), and remove dynamic unbalance at the ends.
If in doubt what I mean, ask. Best regards, Arne

 

[ronfrend]

Just browsing through some old threads and came across this one.

I have had the same problem in the past and it is caused by a very small bowing of the rotor. You need to keep the rotor turning all of the time on a such a sensitive rotor as this. What I did (on an IRD balancing machine) was to slow roll the rotor for about a day before carrying out the balancing procedure. On large rotors you may have to slow roll for a couple of days to get rid of any bow caused by the rotor sagging.

The main thing is not to let the rotor sit for a long time suspended between 2 bearing supports - you wouldn't do it on a steam turbine after a stop - you would keep it turning to avoid a thermally induced bow. If you have to have the rotor sitting for a long time on the supports, install a small drive to slow roll the thing.

ron.frend@predicon.net