Modeling and prediction of surface topography and surface roughness in dual-axis wheel polishing of optical glass

Abstract

Abstract It is rather difficult to achieve uniform surface texture and high convergence rates of surface roughness (CRSR) at the same time in the conventional optical polishing process. The dual-axis wheel polishing (DAWP) technology, which uses a semi-rigid wheel with combined dual-axis rotational movements, shows a strong capability directly changing the rough ground surface status down to mirror status with nanometer level surface roughness, and achieving a uniform and smooth surface texture. In order to provide a better scientific understanding of the surface texture generation in DAWP process, in this paper, two comprehensive mathematical models are proposed. The first model is to predict the generation of micro-topography and roughness within single-axis TIF (tool influence function) area, based on the abrasive wear theory, abrasives distribution investigation and accumulative removal theory. On the basis of this model, another surface micro-topography generation model is also built to predict the micro-topography and roughness generated on workpiece surface in actual DAWP process (both for single and dual-axis modes). A series of fixed spot polishing tests and uniform polishing tests on BK7 glass were carried out to analyze the influence of polishing parameters and motion modes on the generated surface texture and roughness. The measurement results show that the surface micro-topography and roughness within TIF area and on polished surface can be well predicted by the two models. After actual polishing process, the dual-axis mode tends to produce much lower surface roughness with uniform texture. The single-axis mode tends to generate directional surface textures with relatively higher surface roughness. Under the dual-axis mode, the variation of co-rotating speed has less influence on the polishing quality, very high CRSR and uniform texture can be achieved under different co-rotating speeds.