Dear all,

Does anyone have experience of Steam Whirl in the Steam Turbine HIP rotor, what is the signature at the spectrum ?

There is a 350MW Steam Turbine with four Govner valves, the two valves are at the upper side and others are at lower side. Recently, it have a vibration fluctuation problem when the GV changed from throttle to nozzle mode. From attached before and after specturm of HIP rotor bearing, it shown number of broadband low frequenies and below the running speed 50Hz.

Is steam or oil whirl ? Would anyone give me some comments on this matter ?

thanks a lot

kk

spectrum.doc (218 Kb, 67 downloads) Spectrum

Out of curiosity, did the low frequency stuff continue or die away after the transient?

Steam whirl doesn't occur at running speed. This looks like mostly 1X. You should focus in on the lower frequency range. At the low level any steam whirl component doesn't appear that important. Also, steam whirl should have a predominant subsynchronous peak, not broadband.

What type or transducer is this? The type of transducer/measurment may make a difference in the way steam whirl would appear.

Most people look at p-p vibration for displacement.

Thanks for quick reply.
This vibration was taken by proximity probe. Besides, i found this low frequency appearing randomly and varying, and then caused the overall vibration level fluctuated about 10-20 microns pk-pk

thanks..

is it possible caused by vortex problem ?

If the vib event coincides with change in the governor mode (load) it could be steam flow related (not necessarily steam whirl though). Whirl/whip related vibration is usually at a sub-synch frequency close to critical speed (excited by the instability). In your case it seems more like broad-band noise and could be due to unstable steam flow pushing the rotor around under specific load conditions. Not sure what your machine design is but sticky steam valves can cause such thing.

The quoted 10-20 microns change/increase during nozzle mode (load?) change is quite insignificant and overall vibration of less than 50 microns p-p is fine for STG running at 3000 rpm.

BTW, the spectrums say “RMS peak-peak” - shouldn't be either RMS or p-p?

Dear Val_S

thanks for your reply and i am totally agree your view. Since the turbine GV nozzles were just renewed during outage.

In addition, I noticed the fluctuated vibration level would disappeared after the valves above 70% opening and at high load operation.

For the overall vibration level, you are right, our limit was also below 50 microns, Unfortunately, this fluctuation level actually is about 60-80 microns....

Besides, the vibration spectrum was captured from B&K Compass system, i don't know why it show the value such confusing....it should be a microns pk-pk

regards and thanks.

kk

Is the vibration present during part-load operation (<70%)? Was any work done on the seal recently? Is the dominant sub-harmonic peak about 35 Hz? Could you elaborate on the modification (throttle-to-nozzle)? Does it in any way alter the steam inlet angle or pressure? There might have been some change in the damping of the rotor system.

It is an interesting discussion. However, before talking abnout whirl/whip you have to consider if it is tilting or fixed pad bearings. Then you can maybe can exclude something.

Next "Vortex" - come on - it is a steam turbine.

Is the subsynchronous vibration related to the running speed (1/n) and harmonics of it? => indication of parametric excitation.

Is it a broadband excitation between 55% and 75% of the runing speed? => indication of an instability (the first natural frequency is a likely posibility). Bearng wear can change the bearing preload, resulting in less effective dynamic properties (in terms of stiffness and damping)of the bearings supporting the shaft.

It could be flow induced, but then more information and measurements need to be present.

The spectrums are measured in rms and the expressions rms p-p, confusing or not, is the rms value multiplied with the sqrt. of 2, as many people finds this meaningfull.

A good idea is to contact the manufacturer to get a statement or opinion.

This message has been edited. Last edited by: Richard H. Jensen,

I have seen this in numerous steam turbines; it’s not steam whirl, just simply the bearing/s being unloaded by the steam forces. The usual four valve/nozzle arrangement has the two bottom segments admitting steam between about 50% and 75% load with only one of the top valves/nozzle segments. This lifts the shaft in the #1, and possible the #2 bearing, unloading it. Since the 1X amplitude is proportional to the bearing loading (system stiffness) the 1X vibration goes up. You may also see an increase in subsynchronous activity at the same time. This happens regardless of the bearing type. If you can do shaft centerline plots versus load it’s usually fairly obvious.

I know of several steam turbines that must run in full arc/governor valve all the time because of this problem. Changing the sequencing of the control valves/throttle valves can help.

Dear Bob

You are approaching an explanation that very well could be the case. Continuing, this is why steam turbine bearing normally are very heavily pre-loaded, resulting in a relative low system damping (rotor dynamic stability margin) due to the relatively high dynamic stiffness properties of the oil film.
The low system damping is also, why you normally see significant activity below running speed, around the first natural frequency.

Richard