An innovative virtual trial misalignment approach for identification of unbalance, sensor and active magnetic bearing misalignment along with its stiffness parameters in a magnetically levitated flexible rotor system

Abstract

A model-based identification algorithm derived from an innovative virtual trial misalignment (VTM) strategy is developed to identify the rotor unbalance, active magnetic bearings (AMBs) residual misalignment and their displacement and current stiffness parameters. The VTM approach is aligned with the similar concept of trial unbalance utilized in balancing of the rotor. This virtual method for providing the trial misalignment to the rotor is more reliable and effective than the physical trial misalignment, as the method does not need test runs and a precise model of the rotor-sensor-bearing system. The mathematical modelling for the non-contact displacement sensors offset from the rotor center has also been presented and even the offset amounts are identified using this novel approach. To illustrate the proposed methodology, an unbalanced and misaligned flexible rotor system mounted on multiple AMBs has been considered. Non-collinearity between the rotor operating axis and the axis of supported AMBs is the prime cause of misalignment (combination of parallel and angular) fault. A finite element method with a gyroscopic dynamic condensation scheme is employed for acquiring the reduced form of system equations. The numerically generated frequency domain responses are further employed in the identification algorithm for estimating the rotor-AMB system and faults parameters. Besides this, the different levels of the random noise signal as well as rotor modelling errors are added in the developed VTM based methodology to check its efficacy and sensitiveness in a practical rotor model.