Spectral analysis of surface roughness features of a lapped ultraprecision single-point diamond machined surface

Abstract

Single-point diamond turning of soft metals could provide a much shiny surface with an optimal feed rate; however, the machining mark would be left on the machined surface, which caused the roughness cannot be neglected. If the feed rate is too small, the roughness of the surface could be improved but the reflectiveness would be decreased because of damaging the profile. Therefore, it is necessary to develop a lapping method to reduce the roughness by removing the machining mark, while the reflectiveness can be kept at the original level simultaneously. In this study, the novel lapping method, using strands of wool fibers to deliver the abrasive slurry to rub against the lens, was proposed for removing the machining marks on the mold of a lenticular lens by lapping without damaging the profile of the mold. Even though the normal pressure applied by the wool strands onto the mold surface is very low, the coefficient of friction would be increased significantly with the application of the abrasive slurry. The combined effect was to provide a relatively large shear force to lap the surface with a minimal normal force. Therefore, the proposed method could theoretically avoid damaging the lens while effectively removing the irregularities that appeared on the surfaces. In order to evaluate the proposed lapping method, we firstly lapped the machining marks with different lapping parameters (speed, grit size, time, and pressure) to find out the relationship between these parameters and roughness with the same profile of the mold. Secondly, the optimal lapping parameters were designed based on the above lapping results to deduce the best lapping solution for processing the machining marks. Thirdly, the lapped surface profile of the mold was test by optical profiling system, and the features of the surface can be categorized into various spectral distribution groups. Finally, by comparing the variation of the spectral distribution groups, it is verified that based on our proposed methodology, selective removal of surface spectral groups of features becomes possible.