Abstract
Condition monitoring of blades within gas turbines has been and will continue to be of importance in all areas of their use, for maintenance and reliability purposes. Non-intrusive measurement of blade condition is the ambition of most techniques for this endeavour, with a number of methods proposed, investigated and employed for such measurement, with the current dominant method using proximity probes to measure blade arrival time for subsequent processing. It is proposed, however, that the measurement of the casing vibration, due to the aerodynamic-structural interaction within a gas turbine, could provide a means of blade condition monitoring and modal parameter estimation, without requiring perforation of the casing. An analytical model of a gas turbine casing and simulated pressure signal associated with the rotating blades, individual blade vibrations and transfer of stator blade vibrations has been developed in order to understand the complex relationship between casing response and the most important excitation forces.
Due to the force interaction being through a fluid medium, a certain degree of randomness is introduced into the excitations, and the viability of this inherent randomness as a useful aid for separation of the contributing excitation forces from the system response is explored.