On-site high-speed balancing of flexible rotor-bearing system using virtual trial unbalances at slow run

Abstract

An identification algorithm is developed to execute high-speed balancing of the flexible rotor system supported on the conventional bearings, while rotating the system below its critical speed. The system is integrated with Active Magnetic Bearings (AMBs) as a suppression actuator as well as an excitation actuator. The AMB suppresses the vibration of the system and allows to introduce virtual trial unbalances to identify the orientation and magnitude of the residual unbalances present in the system. An Advanced Influence Coefficient Method (AICM) is developed that utilizes the influence coefficients obtained at high speed and unbalances identified at the low speed to effectively estimate the balance masses required for the high-speed flexible rotor balancing. The influence coefficients are required to be obtained just the once, whereas the balance masses can be estimated periodically by identifying the unbalances at low speeds. After balancing, the system can easily cross its critical speeds with less vibration. The AMB also controls abrupt change in vibration amplitude due to any uncertain faults, while operating at high speeds. Additionally, the virtual trial unbalances as a magnetic force generated through AMB reduces the mechanical effort involved in the placing of physical trial unbalances in the rotor system. This method allows on-site condition monitoring of the system periodically to reduce the impairment of high-speed machinery.