Adre measurement: Run-out
A generator is driven by a steam turbine through a gearbox. After a revision, the online measurement system shows increased vibrations at the generator.
By connecting the Adre-box to the present Bently Nevada 3300 rack, the following data was captured:
(click the images for a larger size)
After capturing the steady state values a run-measurement was performed.
The figure below shows the measurement around 300 rpm
We see high slow-roll values:
- Bearing gearbox, generator side: +/- 14 microns -> OK
- Bearing generator, gearbox side: +/- 38 microns -> NOK
- Bearing generator, open side: +/- 25µm -> NOK
This means the vibration value may be influenced by the quality of the measurement position. This measurement position is probably damaged, these damages could be of magnetic or mechanical nature or a combination of both.
The exact value of a slow roll can be obtained by calculation in API541 and API546. Practically, it means to obtain a value of around 10µm.
When we make an approximation of the slow-roll data and vectorially subtracting the measured values, we obtain the following figure:
We thus see that the 1x vibrations are lowering 20µm with the compensation. These values are an approach which indicates the direction, they don’t give an accurate calculation.
The rise of vibrations has happened in the revision. (25 -> 50µm). The run-out measurement shows the cause, an increased slow-roll value.
Possible causes include measurements of magnetic feet, lifting slings, rust, ...
Through this measurement, the cause of the vibration has become clear. Despite the problems of the measurement runs we still have a picture of the vibrations by run-out compensation.
All too often a lot of information is missed because analysis is made on overall values. This case demonstrates clearly the importance of consistent measurements on plain bearings.
Contact ACD NV for the predictive maintenance of your machines!