Self-sensing Techniques of Piezoelectric Actuators in Detecting Unbalance Faults in a Rotating Machine

Abstract

Sensors are inevitable components in a machine or in a mechatronic system, to measure physical quantities. Due to the inherent property of a piezoelectric material, it can be used as an actuator to bring in a displacement in the system by applying an actuator voltage or it can be used as a sensor where a force applied on it is translated as voltage. This property called as self-sensing, is one way to reduce the number of sensors needed in an active system. In the present application, piezoelectric actuators are mounted at the bearings of a rotor. The bearing displacement can be determined from the deflection of the piezos. This deflection can be reconstructed from the current and voltage. By feeding the reconstructed deflection to a finite element (FE) model, faults such as unbalance can be detected. The modal expansion theory helps to determine the deflection at any degree of freedom from few measured signals such as the bearing displacements. Moreover, the forces at each node can be calculated and detected for presence of unbalance faults. With the help of least squares minimization, the magnitude and phase of the unbalance can be determined.