Model-based fatigue crack identification in rotors integrated with active magnetic bearings

Abstract

This paper investigates the behavior of a rotor-bearing system with a breathing crack, under the action of active magnetic bearings, implemented to suppress the vibration induced by the presence of the crack and the unbalance. In such a case, identification of the crack with help of vibration signals may be difficult; other observables such as the controller current may be utilized for the purpose. Linear system equations of motion together with the switching crack excitation function are used to develop the identification algorithm. The use of full spectrum analysis is made to determine the crack force coefficients of the harmonics exciting the rotor in the same and the reverse direction of the rotor spin. Usage of these coefficients in the identification algorithm estimates the viscous damping, disc unbalance and additive crack stiffness. The algorithm has been tested for measurement noise and bias errors in system parameters, for robustness. In practical terms, the situation analyzed in this paper relates to identification of crack and other parameters in a rotor integrated with an active magnetic bearing.