Use this URL to cite or link to this record in EThOS:
Title: A non-linear weighted least squares gas turbine diagnostic approach and multi-fuel performance simulation
Author: Kamunge, Daniel
ISNI:       0000 0004 2700 1599
Awarding Body: Cranfield University
Current Institution: Cranfield University
Date of Award: 2011
Availability of Full Text:
Access from EThOS:
Access from Institution:
The gas turbine which has found numerous applications in Air, Land and Sea applications, as a propulsion system, electricity generator and prime mover, is subject to deterioration of its individual components. In the past, various methodologies have been developed to quantify this deterioration with varying degrees of success. No single method addresses all issues pertaining to gas turbine diagnostics and thus, room for improvement exists. The first part of this research investigates the feasibility of non-linear W eighted Least Squares as a gas turbine component deterioration quantification tool. Two new weighting schemes have been developed to address measurement noise. Four cases have been run to demonstrate the non-linear weighted least squares method, in conjunction with the new weighting schemes. Results demonstrate that the non-linear weighted least squares method effectively addresses measurement noise and quantifies gas path component faults with improved accuracy over its linear counterpart and over methods that do not address measurement noise. Since Gas turbine diagnostics is based on analysis of engine performance at given ambient and power setting conditions; accurate and reliable engine performance modelling and simulation models are essential for meaningful gas turbine diagnostics. The second part of this research therefore sought to develop a multi-fuel and multi-caloric simulation method with the view of improving simulation accuracy. The method developed is based on non-linear interpolation of fuel tables. Fuel tables for Jet-A, UK Natural gas, Kerosene and Diesel were produced. Six case studies were carried out and the results demonstrate that the method has significantly improved accuracy over linear interpolation based methods and methods that assume thermal perfection.
Supervisor: Li, Y. G. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available