Nonlinear dynamic analysis of high speed bearings due to combined localized defects

Abstract

In this paper, the dynamic behavior of a high-speed rotating shaft under the combined localized defects on outer race, inner race and rolling element has been analyzed. An adaptive algorithm based on wavelet packet decomposition and Hilbert transform is used to extract the bearing fault characteristic component from the vibration signals. The reverse Biorthogonal wavelet 5.5 is considered as the most appropriate wavelet for decomposing the bearing vibration signal, on the basis of the maximum energy-to-Shannon entropy ratio criterion. The imperial trend suggested in diagnosing the combined localized defects of rolling bearing is based on experimental analysis at various rotor speeds. The results are presented in the form of envelope analysis, bifurcation diagrams and Poincaré maps that show the appearance nonlinearity in the dynamic response due to the rotor speed and the presence of combined localized defect. Significant peaks are reported at BPFO, BSF, FTF and modulation of BPFI with rotational frequency in envelope spectrum. Fundamental train frequency also shows noticeable excitation. A modified mathematical model shows good agreement with experimental results.