Skip to main content

Guys,

I have a Solar Titan 130 Gas Turbine. It's a 2 Shaft Gas Turbine. 

1st Shaft - Gas Producer Rotor - Consists of Axial Compressor and 2 Stage Gas Turbine. Air Intake End of this rotor is connected through a spline coupling to the accessory drive gearbox. Please see attached layout image.

mceclip0

mceclip1

2nd Shaft - Power Turbine Rotor - 2 Stage Gas Turbine Discs providing mechanical drive to the Centrifugal Gas Compressor

All the radial bearings are fitted with X & Y proximity probes.

Slow Roll response for Bearing 1Y is indicating 30 Microns whereas Slow Roll response for Bearing 1X is 8 Microns. -1X component is present throughout the slow Roll Speed.

mceclip2

Orbits for Bearing 1 are highly elliptical.

mceclip3

Shaft centerline plots show abnormal movement of Shaft Centre within Bearing 1. SCL Plots indicate downward movement of shaft center as soon as the rotor starts turning. Then coming up slowly with increase in rotor speed.

mceclip4

Waterfall and Cascade Plots indicate presence of -1X Component throughout slow roll and operating Speed Range.

Steady state vibration amplitudes are 25 to 30 Microns pp.

What could cause this kind of behavior? Why the high slow roll response is showing up on 1 probe only? I would appreciate any thoughts on the possible reasons for this behavior.

Attachments

Images (6)
  • mceclip0
  • mceclip1
  • mceclip4
  • mceclip2
  • mceclip3
  • mceclip5
Original Post

When inconsistent readings exist in a probe pair I always suggest you start with a β€œloop test” to insure correct wiring.  In a loop test, you get as close as you possibly can to the probe and then perform some action that results in a gap change.  For instance, I might disconnect the extension cable at the proximitor β€œX” Probe of the 3 bearing, then look for a corresponding change in gap voltage at the rack or even all the way back to System 1.  

While doing the loop test, verify that the cabling matches the system characteristics.  For example make sure an extension cable a used with a 5 meter system is not being used with a 9 meter proximitor.

I second John in his recommendation to verify the functionality of the loop.

But I prefer before touching the machine or loop, it is better to have further analysis for the available data. You might discover a need for more actions so you can address them all in the same shutdown window.

The orbit at 335 rpm shows very directional shape along the direction of Probe Y while the reading in the direction of Probe X is very low. An abnormal thing in the Bode plot for Bearing 1 Y probe is that the direct reading is too low, 1 um p-p.

Regards- Ali M. Al-Shurafa

Last edited by Shurafa

Check the axial position and the side that the probes are on, that is a good process for diagnostics.  

Also, one should check the other probe if this is an x-y pair, because you do not know which one may be reading incorrectly.  Too often, people want the measurement that looks good to be correct.  If you have good reason to suspect a particular measurement, it may make sense to only look there - like maintenance was just done on a sensor.

Add Reply

Post
×
×
×
×
Link copied to your clipboard.
×