Texture analysis of rough surfaces using optical Fourier transform

Abstract

An optical Fourier transform technique was developed in conjunction with a machine vision system to study the texture of machined surfaces having a wide range of roughness. Surface-ground samples of different materials were used in the study. The spectra obtained by the optical method were compared with those obtained by the stylus method in terms of spectrum peak and RMS value. Correlation curves were obtained to indicate the relationship between the texture parameters and the average roughness Ra for different materials. Optical roughness correlation curves were also obtained for roughness measurement along the lay direction and subsequently for specimens coated with oil.