Efficient Thermohydrodynamic Radial and Thrust Bearing Modeling for Transient Rotor Simulations

Hydrodynamic bearings are usually employed to support rotating machines, both in the axial as well as in the radial direction. Both bearing types influence the vibration behavior of rotors. Moreover, the oil-temperature influences the hydrodynamic bearing forces. In this work, efficient thermohydrodynamic bearing models for thrust and radial bearings are developed. Run-up simulations are performed for the identification of the influence of the bearings on the nonlinear rotor vibrations. The Reynolds equation, which describes the hydrodynamic pressure distribution in the bearings, is solved using a highly efficient approach. The Global Galerkin approach, using appropriate trial and test functions, is used for the approximation of the solution of the Reynolds equation, leading to heavily reduced simulation times when compared with Finite Difference or Finite Element approaches. For radial bearings, a novel semi-analytical method is developed using also the Global Galerkin approach. The oil-temperature in the thrust bearings is captured through the energy equation, which is decoupled from the Reynolds equation under appropriate assumptions. For the oil-temperature in radial bearings with full- as well as semi-floating rings a global thermal energy balance is used between the two oil-films and the bearing ring. The transient temperature terms in this energy balance are taken into consideration and their significance for the numerical stability of the solver is demonstrated. A turbocharger rotor is modeled in a multibody simulation software. The complete system consists of a flexible shaft, a turbine and a compressor wheel, as well as a thrust bearing and two full-floating ring bearings. The equations of motion of the turbocharger rotor are coupled with the equations of the thermohydrodynamic bearing models and they are solved simultaneously at each time-integration step during a run-up simulation. The simulation results show that the oil-temperature and the gas forces in the axial direction exert a large influence on the rotor vibrations. The geometry of the pads in thrust bearings will be optimized using a novel approach. In this work, statistical and neural network methods are used, avoiding the drawbacks of optimization algorithms. Usually, thrust bearings are optimized for higher load capacity and lower friction losses. Using the proposed optimization approach, thrust bearings can be optimized not only for load capacity and friction losses but also towards a better vibration behavior of the complete rotordynamic system. The validation of the thrust and the radial bearing modeling is performed through comparisons with experimental results. For radial bearings, a standard shaft motion test is used and for the thrust bearing a new testing approach is implemented. The simulation results are in a good agreement with the experimental data