Misalignment modeling and characteristic parameters estimation of a misaligned rotor-bearing assembly: An experimental investigation

Abstract

The present article aims to shed light on rotor misalignment modeling and diagnosis technique. The study comprises three sections. First, it deals with the analytical modeling of a misaligned rotor–bearing assembly beyond the traditional approach. It utilizes Lagrange's equation to acquire the coupled system's mass, stiffness and damping matrices. The inclusion of angular and linear misalignment effect in terms of misalignment matrix to derive system equations of motion is the novelty of the present study. Second, an identification algorithm based on the least squares fit technique is developed to evaluate misalignment and imbalance characteristic parameters. The identification algorithm works on independent forced response data acquired from misaligned rotor-bearing system. Third, using custom-built lab test equipment, the proposed approach is experimentally verified. Under varied misalignment situations, six sets of responses are recorded at each of the four bearing sites at four separate speeds (17, 21, 25 and 31 Hz). The estimated parameters standard deviations are calculated and discovered to be reliable and robust. Under four distinct misalignment conditions (without, pure parallel, pure angular and combined misalignment) the shaft orbit responses at each bearing site are measured and observed that with increasing amounts of misalignment, the amplitude of the orbit response grows.