Statistical analysis of optical Fourier transform patterns for surface texture assessment

Abstract

An optical method for surface texture assessment is developed. This technique utilizes the grey-level histogram of the optical Fourier transform pattern of the surface to derive statistical parameters which are subsequently correlated with the surface roughness as obtained from a stylus instrument. An optical roughness parameter R is obtained for the surface roughness characterization. It is found that it also correlates well with surface roughness for test samples of different materials and also from different machining processes. Optical roughness calibration curves are obtained for the measurement of finishing surfaces with Ra values up to 0.8 mu m. In addition, the method is found capable of detecting the presence of flaws on machined surfaces. This technique provides a fast noncontact method for surface texture assessment of 2D surfaces which could not normally be measured by most other techniques.