Roughness Evaluation of Burnished Topography with a Precise Definition of the S-L Surface

Abstract

Studies of surface topography including processes of measurement and data analysis have an influence on the description of machined parts with their tribological performance. Usually, surface roughness is analysed when a scale-limited (S-L) surface, excluding short (S-) and length (L-) components from the raw measured data, is defined. Errors in the precise definition of the S-L surface can cause the false estimation of detail properties, especially its tribological performance. Errors can arise when the surface contains some burnished details such as oil pockets, dimples, scratches, or, generally, deep or wide features. The validation of proposed methods for S-L surface definition can also affect the accuracy of the ISO 25178 surface topography parameter calculation. It was found that the application of commonly used procedures, available in commercial software (e.g., least-square fitted cylinder element or polynomial planes, regular or robust Gaussian regression, spline, median or fast Fourier transform filters) can be suitable for precise S-L surface definition. However, some additional analyses, based on power spectral densities, autocorrelation function, texture direction graphs, or spectral characterisation, are strongly required. The effect of the definition of the S-L surface on the values of the ISO 25178 parameters was also comprehensively studied. Some proposals of guidance on how to define an appropriate S-L surface with, respectively, an objective evaluation of surface roughness parameters, were also presented.