Orthogonal experiment and analysis of power spectral density on process parameters of pitch tool polishing

Mid to high spatial frequency error (MSFR and HSFR) should be strictly controlled in modern optical systems. Pitch tool polishing (PTP) is an effective ultra-smoothing surface manufacturing method to control MSFR and HSFR. But it is difficult to control because it is affected by a lot of factors. The present paper describes the pitch tool polishing study based on eighteen well-planned orthogonal experiments (OA18 matrix). Five main process factors (abrasive particle size, slurry concentration, pad rotation speed, acidity and polishing time) in pitch tool polishing process were investigated. In this study, power spectral density (PSD) based on Fourier analysis of surface topography data obtained by white light interferometer was used as the results of orthogonal experiments instead of material removal rate and surface roughness. A normalization method of PSD was proposed as the range analysis rule. Three parts of spatial frequency bandwidth were selected and discussed. Acidity is the most important factor in part 1 and slurry concentration is the most significant one in part 2; while acidity is the least influenced one in part 3. The result in each part was explained by two-step material removal mechanism. At last, suggestions in low and high spatial frequency are given for pitch tool polishing.