Cast surface texture characterisation via areal roughness

Abstract

Areal surface texture characterisation leads to significantly enhanced surface information content. To obtain such an improved data set, several methodological steps are required. A suitable filter and nesting index have to be preselected based on surface topography and targeted structure content of the roughness. To select the nesting index, a filter study on synthetically generated surfaces has been performed firstly. Discrete Fourier transform features the selection of the nesting index on basis of manufacturing process affected surface topography. In terms of areal roughness calculation, the nesting index is proposed to be at least two times the present surface period length ls. In our investigations, a nesting index of 8 mm matched best for the analysed sand cast surfaces. As spurious spikes may occur by overexposure during image acquisition, a procedure is presented to minimize these by invoking Hampel filter based strategies. The newly introduced sub-area evaluation methodology further improves the roughness-based information content as distinctive, local surface structure characteristics can be traced back to corresponding areal roughness parameters. Finally, analysis of sub-area roughness parameters reveal that the arithmetical mean height Sa is log-logistic distributed.