Computational Study of the Non-Parametric Squeeze Film Damper Bearing Inverse Model Based on Artificial Neural Networks Applied to a Rotor-Casing System Running on Unsupported SFDs

Abstract

Squeeze Film Damper (SFD) bearings play a vital role in attenuating large amplitudes of vibration due to their relatively simple assembly in aero engine designs. The modern aero-engine structures, typically, have at least two nested rotors mounted within a flexible casing via squeeze-film damper (SFD) bearings. There is a growing body of research into identification techniques for bearing models for use in rotor-bearing analysis to improve reliability and/or efficiency of implementation. The authors’ previous work has shown that, for cases where there is no adequate linear connection between the rotor and casing, the identification of the unbalance from externally mounted sensors requires a virtual instrument that can determine the vibration of the rotor relative to the casing, as a substitute for internal instrumentation. The present study is devoted to determining the effectiveness of the inverse SFD model (under different unbalance state conditions), when it is applied to a rotor-casing system, wherein the rotor runs on two unsprung SFD bearings. The validation of the inverse SFD model enables its use in a future study of the identification of unbalance in such complex systems.