Analysis And Optimal Design Of High Speed Hybrid Ceramic Ball Bearing

High speed is one of the most important requirements of mechanical products development. To realize this requirement, high speed ball bearing is needed to develop. In the operation condition of high speed, the speed of steel ball bearing (SBB) is limited by the rapidly increased centrifugal force and heat generation. This limitation may be overcome by a hybrid ceramic ball bearing (HCBB) which utilizes ceramic balls instead of steel balls.The performance of the HCBB is quite different from that of SBB in the operation condition of high speed, since the material properties of ceramic is different from that of steel. The conventional fatigue life and temperature rise analysis of bearing is based on empirical theories which are established on a large amount of historical experiment data. However, these empirical theories are not adaptive for ceramic material and high speed condition. Besides, the conventional design method which is for SBB may not exert the superiority of HCBB.In this paper, Ioannides-Harris theory is applied to the fatigue life calculation of HCBB. Finite element method (FEM) with submodel technology is adopted to solve stress field of ball-race contact subsurface which is needed by the Ioannides-Harris theory. Then, the calculation of fatigue life is simplified by a two-dimensional interpolation method.After that, the high speed HCBB is simulated by FEM to obtain dynamic parameters which is needed by frictional power loss calculation. Then, the frictional power loss is applied to a temperature rise finite element model. Results show that load and shaft speed have great effect on ball-inner race spin frictional power loss, but little on ball-outer race spin frictional power loss and ball-outer race differential slip frictional power loss. The load and shaft speed have great effect on temperature rise of inner race and ball, but little on that of outer race.Next, a multi-objective optimal design model of high speed HCBB is proposed. The objectives are the fatigue life, the spin frictional power loss and the axial stiffness. An algorithm named non-dominated sort genetic algorithm-II (NSGA-II) is adopted to solve this multi-objective optimization problem, since the objective functions of HCBB are implicit. During solving the optimization problem, a self-adaptive method is developed to assign initial value to a set of nonlinear equations.Finally, four types of HCBB are selected as examples to be optimized. The Pareto optimal front and the distribution of design parameter are analyzed. The current and optimal designs'variation of objective value with operation condition is studied. Results show that with the same fatigue life, the spin frictional power loss of optimal design is equal to or smaller than that of current design. With the same spin frictional power loss, the fatigue life of the optimal design is longer or equal to that of current design. The four type bearings'axial stiffness of optimal design are all larger than that of current design.