Uncertainty quantification and probabilistic reliability analysis for the self-excited vibration of a spline-shafting system

Floating involute splines are widely used in large-torque, high-speed aerospace power transmission systems. The spline rotor system may suffer from self-excited vibration due to poor lubrication. The unavoidable manufacturing tolerances, mounting errors, component deformation, and other factors make the self-excited vibration occurrence conditions have significant uncertainties. This paper provides new insights into the uncertainty quantification and reliability analysis of self-excited vibration of a spline-shafting system based on an advanced Kriging surrogate model. First, a rotor nonlinear dynamics model considering tooth surface friction is established. Then an effective Kriging surrogate model is developed to quantify the uncertainty propagation in nonlinear rotor dynamics considering the randomness of parameters such as friction coefficient, and load torque, which enables the rapid estimation of statistical information of the self-excited vibration response by the time-domain numerical integration method. The probability of instability and reliability of the system under multi-parameter uncertainty for the occurrence of self-excited vibration is also predicted. Finally, the validity of the proposed method is verified by Monte Carlo simulation. This contribution will further enrich the theory and applications of probabilistic statistical analysis and reliability assessment of complex rotating machinery.