Nonlinear dynamic analysis of full-ceramic bearing-rotor systems considering dynamic waviness variations

During the bearing operation, surface waviness can cause severe nonlinear vibrations and reduce the life and reliability of bearing-rotor systems. A temperature increase can deform the surface waviness, which can exacerbate this unstable vibration. To address this problem, a dynamic waviness model considering thermal deformation is proposed, and the dynamic support stiffness is calculated and introduced into the dynamic model of a 12-DOF full-ceramic bearing-rotor system. The Newton–Raphson and Newmark-β nested iterative solution method combines the quasi-static and dynamical models. The bifurcation, maximum Lyapunov exponent, and Poincaré mapping analysis methods serve to analyse how thermal deformation, waviness amplitude and other parameters affect the system nonlinear vibration. Experimental measurements are conducted to verify the model accuracy and reveal that a larger waviness amplitude causes a delayed motion state and expands the influence of the thermal deformation. The wavenumber is close to an integer multiple of the ball, and the time-varying displacement excitation curve shows a monotonic variation trend, which makes the system violently vibrate. The model effectively reveals the characteristics of the failure frequency and provides important theoretical support for the fault detection of full-ceramic bearings.