The method of adaptive modelling of Gas Turbine performance, proposed in the past by the present authors, is finding a new application in component fault detection. An appropriate choice of component characteristic parameters gives the possibility of characterizing their “health condition”, by appropriate processing of selected measured quantities according to the principles of adaptive modelling. Both a general deterioration of components and identification (kind, location) of component faults can be achieved. The proposed methodology is presented together with an experimental investigation of Gas Turbine faults. A commercial Gas Turbine has been used as a test vehicle and various kinds of faults have been implanted. An analysis of the results of the investigation provides information about the influence of the occurrence of such faults on engine performance. On the other hand, these data provide a basis for demonstrating the capabilities of the above mentioned methodology, as a powerful tool for diagnostic applications.

This content is only available via PDF.