Loading...

Model based compressor fault identification using stage stacking technique and nonlinear diagnostic systems

Hosseini, H. R ; Sharif University of Technology | 2008

448 Viewed
  1. Type of Document: Article
  2. DOI: 10.1115/GT2008-51032
  3. Publisher: 2008
  4. Abstract:
  5. Compressor fault identification is an important part of gas turbine diagnostic systems. Model based techniques have been used widely in this field. In this paper the performance of two compressors has been simulated using stage stacking method. One of them is the NASA 10 stage constant tip diameter compressor. The other one is the Siemens V94.2 engine's compressor. Then a new and more real compressor fault model is introduced and the effect of different faults on compressor performance has been studied. In this paper an intelligent fault diagnostic system has been developed which is able to detect different faults of the compressor. In all of the cases of faults, degradation in inlet flow coefficient of the compressor has been considered besides degradation of pressure rise coefficient and stages' efficiencies, which has not been considered in any available study. It is shown that the fault identification system is very powerful in detecting single type faults. In the case of combined faults it was found that the faults which have been distributed in more than two stages will be identified. Copyright © 2008 by ASME
  6. Keywords:
  7. Compressor performance ; Diagnostic systems ; Fault diagnostic systems ; Fault identifications ; Fault model ; Gas turbine diagnostics ; Inlet flow ; Model based techniques ; Model-based ; Pressure rise coefficient ; Siemens ; Stage stacking ; Two stage ; Degradation ; Gas turbines ; NASA ; Compressors
  8. Source: Proceedings of the ASME Turbo Expo, 9 June 2008 through 13 June 2008, Berlin ; Volume 2 , 2008 , Pages 213-221 ; 9780791843123 (ISBN)
  9. URL: https://asmedigitalcollection.asme.org/GT/proceedings-abstract/GT2008/43123/213/328734