Hi ,

I recently took vibration measurements during run-up of a 50Hz 100MW Toshiba Steam Turbine. Recently this turbine has lost it's DC supply and all the bearing has been damaged.
The bearing has been changed but the bearing clearance has slightly exceeded it's design clearance by 10 micr. The vibration is near to the alarm limit and what may be the cause of this vibration characteristic.

HP_Turbine_Brg_1.doc (163 Kb, 62 downloads)

Tajul,

Please post previous and current run-up waterfall plots? Also are those alarms reasonable considering generally recommended R/C ratio?

David

Tajul,
It is evident that microns of loosenes in bearing means a big issue for a large machine like this, what is you have to do is perform a run test, increase the load till max.load recording the following paremeters

Time - Load - B1 Ampl.(mm/sec)- phase - B2 - phase - speed

Please post your data to check it out.

Unbalnce will leave a nearly constanst level and phase, you know that bearing loosness tends to reveal unblance.

It is possible that an open bearing clearance can result in higher vibration. Did you also change train alignment?

Does this vibration bother your? Do you have time for corrective action? Do you have a new bearing? Changing a bearing may not correct this to the level (what level?) you desire.

Tracking the 1X (and ensuring not other odd features arise) as already suggested would be good, if you can afford this. Hopefully, you have tracked the 1x over time (and not 1X stuff). You should pay attention to all bearings.

If the spectra plots are an indicator, the waterfall plots may be quite boring, which is good.

A balance shot can be quicker and more economic than replacing a bearing. Replace the bearing when you can afford the down time and have the parts.

I do not support the bearing repalcement, it is a high cost in capital and down time,balance will be moreeconomic, the matter is that the machine after a balance will be more sensetive to unbalance, you can live with that.

Please post a bode plot with 1x data. This plot will answer a couple of questions. Is the imbalance journal run-out or a rotor rub induced? Also how does the shaft center line movements differ from past history. B1 bearing appears to move half as much and more typical for a sleeve type bearing then B2 bearing. Are bearings sleeve or tilt pad?
Could B1 bearing be preloaded causing the rotor to rub? Something happened in the shaft centerline plots around 500 - 1000rpm that needs to be identified with B1 bearing showing more of this event.

Also do you have phased case readings to figure out the case to shaft relative motion.

TWFs at steady state will be helpful. Also transient orbits may tell a lot. Troublesome IMO, in addition to Eric's comments, is excessive altitude angle for both bearing, more for bearing B2.
Is slow run data available to assess runout?
Is 330 micron radial clearance normal for this turbine?