Blades rubs and looseness detection in gas turbines - operational field experience and laboratory study

Operational experience showed that blade related failures to be common faults in gas turbines. A review of detection and assessment techniques for blade related failures in gas turbines are presented. This paper examines the use of vibration analysis and monitoring of blade pass frequencies and its...

Full description

Saved in:
Bibliographic Details
Main Authors: Leong, Mohd. Salman, Lim, Meng Hee
Format: Article
Published: Vibroengineering 2013
Subjects:
Online Access:http://eprints.utm.my/id/eprint/49997/
https://www.jvejournals.com/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Operational experience showed that blade related failures to be common faults in gas turbines. A review of detection and assessment techniques for blade related failures in gas turbines are presented. This paper examines the use of vibration analysis and monitoring of blade pass frequencies and its side bands for blade rubs and cracks, and wavelet analysis for blades looseness detection. Case studies of operational gas turbines used in power generation are presented. A case of severe rub from a cracked shaft (that exhibited increased vibration trend over time) showed severe harmonics of running speed during a full rub. A longitudinal crack was subsequently found on the rotor shaft. Another case study demonstrated significant increase in amplitudes associated with the blades passing frequencies with increased side band activities from stator blades and labyrinth rubs prior to shaft seizure. Actual field data and historical comparisons of vibration spectra and wavelet maps of the gas turbine are presented. Looseness in the packing pieces of blade roots are also potential problems in some gas turbine designs which are difficult to detect under field conditions. The use of wavelet analysis for simulated blades looseness in a laboratory study showed changes in wavelet maps for rotor coast down measurements suggesting potential in detection for blades looseness. Changes could not be detected from FFT spectra, coast down spectra nor were it detectable from steady state FFT and wavelet analysis.