Identification of Multiple Faults With Incomplete Response Measurements in Rotor-Bearing-Coupling Systems

Abstract

The turbo-generator system of the contemporary rotating machinery consists of the driver and driven shafts, which are coupled through either a rigid or flexible coupling. It needs precise and steadfast prediction of its dynamic attributes. The multiple–faults algorithm developed in [1] that identifies the bearing and coupling dynamic parameters along with residual unbalances in a rigid-rotor and flexible-bearing-coupling system. In actual practice the shaft would be flexible with MDOFs, and it would be difficult to measure all DOFs due to practical limitation of measurement locations and number of sensors. To overcome this in the present paper a dynamic condensation scheme has been applied to condense all angular DOFs (except at coupling location) and several linear DOFs (other than the bearing, disc and coupling locations). Numerical experiments have been performed to illustrate the proposed algorithm. The measurement noise has been added in numerically simulated responses and is used in the present algorithm to identify the parameters. The algorithm is found to be robust with different level of measurement noise. Modelling errors are also considered of few physical parameters and estimate shows the marginal robustness of the algorithm with the changes in the physical parameters.