Variable asymmetric morphological profile filter for roughness analysis

Abstract

Morphological profile filters effectively complement linear and robust Gaussian regression filters in roughness analysis. The advantages of morphological filters include simple elimination of edge effects and no need to exclude a form component from the profile in advance. The standard ISO 16610-41:2015, morphological profile filter is nonlinear and naturally robust. Therefore, it can be used to analyze multifunctional surfaces, for example, after honing. However, due to the nonlinearity of the filter, the obtained result strongly depends on the selected value of a nesting index and cannot always be compared with other filtering methods. The alternating symmetric filter according to ISO 16610-49:2015 provides additional tuning options. By sequentially combine openings and closings operations, it is possible to obtain different degrees of suppression of profile valleys and peaks. The most promising would be asymmetric morphological filter, which uses different nesting indexes for combinations of opening (O) and closing (C) operations. The present paper investigates the influence of the nesting index on the height roughness parameters. For this purpose, profile with preferential depressions after honing was analyzed. It was found that only two combinations of alternating morphological filter give different results, exhausting the variety of morphological combinations of opening and closing. In this case, the degree of suppression of valleys and peaks depends on the form of the primary profile. The filter m = C(O) proved to be the most effective. At that, a smaller nesting index should be taken at first, and then a larger one.