Identification in a Magnetically Levitated Rigid Rotor System Integrated with Misaligned Sensors and Active Magnetic Bearings

Misalignment is one of the serious faults that occur in rotating machinery. This fault may be caused by the offset between the axis of the supported shaft and the bearings axis or the offset between two coupled shafts. Excessive vibration can be developed in the machines due to the misalignment fault, which can result in the breakdown of the complete machine. Therefore, there is a need to explore and analyze the vibrational behavior of a faulty rotor-bearing system and identify the fault for the untroubled performance of machines.

An innovative virtual trial misalignment approach based on trial bias current is proposed to study the dynamics of an unbalanced rigid rotor misaligned with the supported active magnetic bearings and eddy current proximity displacement sensors. The same approach has been also utilized to identify the residual offset amounts of displacement sensors located at active magnetic bearing locations.

Numerical results are obtained by solving the system’s equations of motion on SIMULINKTM platform. The results demonstrate that the displacement and current signals are sinusoidal in nature due to the presence of unbalanced force. Moreover, the size of orbital responses gets enhanced with increments in the misalignment level and noise signal errors. The virtual trial misalignment strategy could also identify the offset amounts of sensors.

This paper proposes a virtual trial misalignment approach to analyze the effects of unbalance and AMBs as well as sensor residual misalignments on rotating machinery integrated with active magnetic bearings. For executing this motive, a rigid rotor linked with two discs (at offset positions) mounted on two misaligned AMBs at the end locations is considered and mathematically modeled. The dynamic effect of AMB and sensor residual misalignment on the vibrating nature of the rotor is presented for different misalignment levels, at a single spin speed of the rotor as well as ramp-up speed. The residual misalignments of eddy current proximity sensors are also identified with the help of mathematical modeling of misaligned sensors and virtual trial misalignments of the rotor. The proposed method is found to be more reliable and efficient as compared to traditional approaches for the identification of faults.