During transient measuremts on a steam turbine the rotor rubbed only during shutdown at its critacal speed. The direct unfiltered orbit shows a dwell time of 23ms (38% of one rad./s) this 23 ms was parrallel with a straigt line. Oh yeah, the 0,5X reversed component was 80microns...keyphasor dots where 180 dergrees from eachother.

The next start-up reveals no rub related vibs. We tripped again and suddenly the 0,5X component was like a cirkel (!) and shows also keyphasor dots 180 degrees cross each other. the direct unfiltered orbit amplitude at that time was 200microns.... This was shown on both the bearing positions!! The shaft centerlines are laying in the middel on both bearings.

Question:
The main question is : Why are these orbits looked different at the same speed (4100RPM)? the first orbit had straight dwell time(23ms), the second was like a smooth big cirkel??
(ps. the direct slow roll direct data was 14micron @ both bearings, no scratches where seen in TWF)

You are asking

quote:

Why are these orbits looked different at the same speed

. How about posting some data (graphics) for us to examine?

John from PA

I think the keyphasor dots are not exactly 180 degree apart . So the frequency is <0.5X. May be whirling.
Irshad

AKHTAR,

New knowledge comes to new insights..
Studiing the data even more because of a case historie which was written in the orbit i dealed with the backflow fenomena.. The rotor rotation became ccw and was then whirling. The was bicking me al the time like a splinter in my mind. The root cause is probarly a bad device of a compressor valve. The gas was acting like ramp from 8800RPM CW till 0 (within 5 seconds) and reaccelrated to 4400RPM CCW. The runout was then exponential. Because the configuration was set to delta rpm of 17RPM the last data wasn't captured due to overkill proparly. I know this is serious matter due to safety and reliability.
Does anyone know how to make calculations due to stress on the couplings en rotors???

(see for more info : http://maintenanceforums.com/e...51015103/m/842106763 )

This message has been edited. Last edited by: monitor,