After overhaul, from a 8 MW Steam Turbine Inlet bearing, the enclosed Toggle circular waveform was received. The measurement was taken from Bently Nevada Monitor BNC output, tach signal from keyphaser.

One of the blade was broken at the last stage exhaust end. The user did not replace it. The turbine was balanced at a balance machine. With this the turbine can pass the critical speed easly but could not reach to the turning speed 7500 RPM. The field balance calculate the weights 180 degree away from the balance machine results. This time sometimes it can pass the critical speed, sometimes not. The monitoring system shutdown the turbine.
Ones, When the critical speed was passed, the vibration readings are good at no load. After 30 minutes, start to get load 4 MW is OK. When reach to 5.8 MW the vibration values get high and the Turbine shut down.
The graph was taken at no load condition.
The time waveform shows some spikes.

The main oil pump (vertical) is Turning at 1500 RPM. The vibration levels are low at pump. But at pipes huge signal was taken at 2x1500= 3000 RPM. At Turbine bearings no indications of this frequency peak. But at gear box yes.
This is close to critical speed of Turbine wgich is written between 2500-3500 RPM. With Bode Graph it is 2887 RPM.
This Oil Pump stop when the Turbine reach to 7500 RPM

I have just this limited info. If any experience and idea for the shape of toggle waveform, please advise.

Spectrum and waveform for the same measurement

Graphs with other format if can not seen

ST.rtf (170 Kb, 23 downloads)

quote:

One of the blade was broken at the last stage exhaust end. The user did not replace it. The turbine was balanced at a balance machine.

Did they cut the blade off opposite the one that was broken as well as machine off the broken portion? If not one could see this being load sensitive.

Is the plot of shaft displacement or casing? Shaft rider or prox probe? Looks like there is a little runout in the time domain plot, and the amplitudes don't look too bad, either if this is shaft relative or shaft absolute.

A field balance of a turbine passing a critical speed should include the influences of the critical speed as well as running speed and any significant load effects.

Bill

No they did not cut the other side blade.
Since it is 59 blade, they need to cut 2 blade.
So next they need to buy 3 blades, where the cost is 3x20,000 Pound.

So, with your dision, this is a mistake.!!

The measurements were taken proximity probes. Shaft displacement. Not from casing.

I'm not a turbine expert, but with one blade missing and not balancing them out sounds like problems.

Back in 1991, I had the priviledge to see a turbine rotor do a loop some 130 feet in the air. It was estimated to be spinning at 10,000 rpm before escaping through the roof.

I called Leroy to see this one cause he never saw a wreck like that!!!!!!!!!

Gary B

You didn't show any data from the high vibration. Was it 1X? If the vibration gets high on another component near 3000 cpm, it could be an instability. Too many open areas, these need to be resolved with data.

Can you get polar plots if it is a 1X problem? If it is not a 1X problem, balance isn't the main problem. The missing blade will cause a 1X imbalance depending upon the flow past this stage.

There could also be a some rubbing on one side. The time waveform seems to be normal (at the top) and bump (at the bottom)of teh waveform. The orbit plot shows the same thing. It could be an internal misalignment (seal, etc) or due to imbalance.

I would suggest you have a lousy shaft surface that the probe is looking at. It appears highly scratched up -- or might it have been chromed?

The amplitude of vibration (7 microns)does not indicate a problem. Before I would read too much into the orbit, I would like a precision ground and burnished probe surface for the probe to look at.

tm

oops, I meant 14 microns.

tm

Without seeing the slowroll it is difficult to say if this has a small rub or some runout. However, I doubt that the runout is much over 1/2 the signal; although it may be. The runout could be smaller.

With only 16.7 micrometres pp the runout may be in the 6 to 8 micrometre pp range,which needs to be clarified. if the runout is 6 micrometres pp or less this would be acceptable, 8 would be slightly higher than desired. All in all, the runout may not be excessive.