Optical Sensing of Surface Roughness for Polishing Robot. (3rd Report). In Case of Anisotropic Surface Roughness.

Abstract

For the reasons of harsh environment in polishing procedures, an optical, non-contact measurement of surface roughness is better suited to in-process measurement than contact sensors such as a stylus meter. The authors have proposed a new optical method for measuring surface roughness by use of a halogen light source and a CCD camera. Based on the Torrance-Sparrow model, the proposed procedure theoretically converts the data of the reflected light into surface roughness represented in Rq (roughness in root mean square).In this paper, we attempt to apply the proposed method to estimate a ground direction. From the results of the computer simulation and the experiments, the ground direction can be estimated within the error of 10 degrees. Further, the root-mean-square roughness can be evaluated within the error of 5% to the true value.