Mechanical rotor unbalance monitoring based on system identification and signal processing approaches

Mechanical unbalance is an important source of vibrations that can cause malfunctions in rotodynamic machinery. In industrial applications, unbalance is a critical issue for mass production machines. Previous studies for detecting and monitoring unbalance are based on balancing machines, trial weights, and intrusive actuators, while other studies rely on signal processing techniques, finite element analysis and physical modeling. These methodologies have some critical drawbacks, especially when non-intrusive monitoring is required, such as having to trial weights or determine constructive parameters such as mass and stiffness. The proposed approach is based on detecting and monitoring the unbalance condition in rotatory machines using data extracted from vibration sensors and a rotation sensor fitted to the system supports. The methodology comprises two main steps: identifying the appropriate speed range for unbalance monitoring and the modal parameters of the rotor, and determining and continuously monitoring the unbalance condition. Signal processing and system identification techniques are used to estimate unbalance in the rotatory machine. Experimental results for two rotodynamic systems demonstrate satisfactory performance in identifying and monitoring different unbalance conditions.

Automated summary

Mechanical unbalance is a significant cause of rotodynamic machinery malfunctions. Previous studies had drawbacks in detecting and monitoring unbalance, while the proposed approach utilizes data from vibration sensors and a rotation sensor to effectively detect and monitor unbalance conditions.