When monitoring vibration on large gas or steam turbines and generators with fluid-film bearings, the relative movement of the shaft within the bearing is typically measured by a pair of shaft proximity (eddy current) transducers. Data from these transducers is used to produce a variety of plots on a condition monitoring system including orbit displays and shaft centreline plots.
Typically these plots assume that the transducers are fitted in the vertical and horizontal direction at the bearing. However, in most practical situations this will not be possible, due to pipework or because the bearing casing is split horizontally. In these case they will be fitted at either 45 degrees from the top or the bottom of the bearing. The important fact is that the probes remain perpendicular.
Note The general convention for the naming or identification of these transducers goes back to using a dual-channel oscilloscope to view the combined XY or ‘orbit’ display from the resultant time history traces.
When viewed from the ‘driven’ end of the bearing and for probes fitted to the top of the bearing the probe on the left-hand side of the bearing is named Y and the X is on the right. This is irrespective of the direction of rotation.
The diagram to the left shows typical transducer mounting positions and the rotation correction required to return the display to true Vertical and Horizontal. Following rotation correction the Y sensor becomes vertical and the X horizontal.
Shaft proximity probes are designed to measure along its axis only and so do not measure anything in the perpendicular direction. If the two probes were not mounted 90 degrees apart then the X and Y measurements would not be independent and any subsequent orbit plot would be skewed.
The following example shows why it is important to understand the location of the transducers when viewing vibration data. If we look at a typical Trend display from PROTOR for the variation in the 1st order amplitude and phase for a pair of shaft displacement probes against speed for a rundown (coastdown) of a Steam Turbine we can see how the orientation of the transducer affects the data we see.
In the following display (Figure 2), the blue curve is for the Y- transducer and the red curve for the X-transducer for a bearing.
If we focus on the amplitude and frequency of the main peak or resonance in the curve (Figure 3) then we can determine that the maximum amplitude for the Y transducer appears at 1640 RPM and has an amplitude of 180 um pk-pk.
Note: In the latest version of PROTOR the cursor window now has ‘Next’ and ‘Prev’ options, these move the cursor to the next or previous data scan and can be used to simply step through the measured data to find the point you want to identfy.
We might assume from this that the maximum vibration amplitude experienced at the bearing would be 180 um pk-pk. However we must remember that this measurement is taken along the axis of the transducer only.
In PROTOR we are able to define the location of the pair of transducers and to then combine these into the equivalent Vertical and Horizontal measurements. If we now select the transducer orientation option to be enabled the resultant trend curves change to that shown in Figure 4.
On applying the orientation translation, where the Y-Probe transforms to the vertical direction and the X-Probe to horizontal, we see that the peak amplitude in the vertical direction is now 189 um pk-pk at 1670 RPM. That is, the vibration amplitude in the true vertical direction is actually higher and at a slightly different frequency to that measured directly by the Y-probe. This identifies the true vertical critical speed for this bearing.
This can also be seen by viewing the Orbit display for the pair of the transducers. As mentioned above the Orbit is the display of the two time histories collected for the pair of transducers, one represents movement in the Y-direction and the other in the X-direction. This orbit represents the movement of the centreline of the shaft within the bearing for one or more cycles. Figure 5 shows the 1st order ‘filtered’ orbit. Initially this is shown without any orientation correction applied, that is the vertical axis on the graph represents the vibration vector along the direction of the Y transducer.
If we now apply the orientation correction (Figure 6), the y-axis on the graph now represents the true vertical vibration and the x-axis the horizontal and the orbit has been rotated by 45 degrees.
With the Orbit display now showing the effective Vertical and Horizontal vibration and if we now step through the orbit displays for the speeds 1670 to 1640 RPM (Figure 7) we see how the orbit processes in the anti-clockwise direction as the speed decreases. At 1670RPM the orbit is at a maximum at the vertical direction but as the speed decreases the orbit major axis moves towards the axis of the Y-transducer until it reaches a maximum in this direction at 1640 RPM.
In conclusion, when viewing vibration data from a pair of shaft displacement probes always be aware of where the transducers are fitted and the conventions used for naming or labelling the signals from the transducers.
Don Davies graduated from the Institute of Sound and Vibration Research (ISVR) at Southampton University in 1979. Don specialises in the capture and analysis of vibration data from rotating machines such as power station turbine generators. He developed and is the Product Manager for PROTOR. Don is a member of the British Computer Society.