Improved magnetorheological finishing of a 3D K9 glass surface for ultra-precision optical applications

This study explores the improved magnetorheological finishing (MRF) process for complex 3D K9 glass surfaces by minimizing non-uniform working gaps. MRF process parameters and MRP fluid composition were optimized using response surface methodology and validated via an artificial neural network, achieving prediction errors below 5%. The optimized process reduced surface roughness from 18.46 to 1.09 nm and improved transmittance from 75.86% to 84.39%. Form error was also minimized while preserving dimensional accuracy. These results demonstrate the process's suitability for advanced optical applications, including high-power laser systems, photolithography optics, space optics, and precision photonic components.