Study of surface texture measurement errors

Abstract

The textures of the surfaces of machine elements affect their functional properties. It is practically impossible to assess the surface texture using only one parameter. However, the number of parameters should not be excessively high. The selected parameters should be functionally relevant, manufacture related and of small sensitivities to measurement errors. In this paper, a number of errors in various machined surface textures measurements were analyzed. Parameters sensitivities to mechanical filtration by a stylus tip, presences of high-frequency noise and spikes, stitching, change of measurement method, sampling interval increase, change of sampling area, and digital filtration were studied. Robust parameters and those that were susceptible to errors were obtained. The spatial parameters were found to be the most robust, whereas those describing the shape of the height distribution, Ssk and Sku, were highly sensitive to measurement errors.