Bearing Looseness Fault Characteristics Analysis of Rotor System Based on Incremental Harmonic Balance Method

Abstract

Support or pedestal looseness fault exists in mechanical system typically.A new dynamical model of an eccentric motor rotor system with a piecewise-linear stiffness was proposed, and the incremental harmonic balance method was employed to obtain analytic solutions of the model. Then, the time domain waveform and amplitude spectrogram were achieved by applying the analytic solutions to program and iterate in the MATLAB software. Finally, the characteristics of looseness was analyzed. Results show that this research method is feasible, and this paper may provide a theoretical basis for the solution of complex bearing looseness model analytically and the diagnosis of its fault.

The Proposal and Research on the Most Basic Problem of Fault Diagnosis

“…Liu Yang et al [61] established a three support loose rub impact coupling fault double disk rotor bearing system model by using the finite element method. Based on numerical integration method and analytical approximate solution, such as Chu Fulei et al [38], Li Zhenping et al [62], Li Hongkun et al [63], Yang Yongfeng et al [64] Changli et al [65], Liu Xiandong et al [51], Chen Guo [66], MAH, Zhao x, Teng y, et al [67], Xiong et al [68] were solved by numerical methods such as Runge Kutta method, Newmark-β-method, shooting method, progressive method and incremental harmonic balance method.…”

A Rotor Looseness‒Rub-Impact‒Resonance Fault Mechanism Modeling

“…Based on numerical integration method and analytic approximate method, such as Chu Fulei et al, Li Zhenping et al [29], Li Hongkun et al [30], Yang Yongfeng et al [31], Liu Changli et al [32],Chen Guo [33] MA H, ZHAO X, TENG Y, et al [34], Xiong et al [35] used Runge-Kutta method, Newmark-β method, shooting method, incremental method and incremental harmonic balance method [6,[29][30][31][32][33][34][35]. Numerical solution method generally needs the condition to be simplified, so objectively produces the disparity with the actual state; The numerical solution method needs to assign a value to the model, for this reason that this general model becomes the specific sampling, the solution obtained is equivalent to the experiment result.…”