Co-optimization of magnetic abrasive finishing behaviors of zirconium tube surfaces with Fe-6.5 wt% Si/SiC abrasives using BP neural network and response surface methodology

Abstract

A novel abrasive of Fe-6.5 wt%Si/SiC composite was prepared for magnetic abrasive finishing (MAF) to achieve fine machining of zirconium alloy tubes. The optimized parameters of magnetic pole speed R= 16.0 r·min−1, tube speed Q= 1220.5 r·min−1, abrasive quality P = 119.3 g, and machining gap N = 2.30 mm were obtained by response surface analysis. The BP neural network model was used to predict the optimal machining effect of MAF. It showed that the predicted results of BP neural network model are quite close to the actual test results, with an error of 7.526%. Finally, the roughness of zirconium alloy tube was reduced from 0.207 µm to 0.113 µm, resulting in a 45.41% improvement rate. Fe-6.5 wt% Si/SiC composite abrasive was an efficient magnetic abrasive to improve the surface quality of zirconium alloy parts.