Two-objective process optimization for grinding of outer ring of silicon nitride ceramic bearing

The machining quality of ceramic bearing rings plays an important role on the rotation accuracy and service performance of the bearing. Firstly, based on a large number of cylindrical grinding experiments, the unitary model of surface roughness with respect to process parameters and the unitary model of raceway roundness with respect to process parameters were established by the least square method. Secondly, on the basis of the unary model, the particle swarm optimization(PSO) algorithm was used to establish a multi-element model of surface roughness with respect to process parameters and a multi-element model of raceway roundness with respect to process parameters. Finally, the surface roughness and raceway roundness are optimized by PSO algorithm. So as to explore the optimal processing parameters of bearing outer ring. The results show that the relative error between the predicted value of the multiple composite model of the surface roughness with respect to the process parameters and the actual processing value is between 5.83% and 8.99%. The relative error between the predicted value of the multivariate composite model of the raceway roundness with respect to the process parameters and the actual processing value is between 4.62% and 8.01%. The process parameters obtained by the dual objective function optimization are the grinding wheel speed of 56.0 m/s, the feed rate of 0.012 mm/min, and the workpiece rotation speed of 215 r/min. Research shows that the multivariate model can predict the actual processing conditions more accurately. The roughness and roundness values under the optimal process parameters are 0.130 μm and 2.20 μm, respectively. Compared with other parameters, it can simultaneously ensure that the roughness and roundness are small.