Summit identification of anisotropic surface texture and directionality assessment based on asperity tip geometry

摘要

Areal topographic data of engineering surfaces may be expressed by a series of discrete sampling points in the horizontal plane, usually at orthogonal crossing points. A set of the areal data is likely to include high frequency irregular components. This leads to difficulty in identifying asperity summits or pit bottoms on the original measured data. A certain surface filter may be used to eliminate those noise-like components. It has been pointed out for a long time that the definitions of the summit and bottom, as well as local slope of facet, are ambiguous and not generalized [1]. This is especially the case with anisotropic or multi-directional surface textures. A data processing method called "four or eight nearest neighbors" [2] has been widely adopted to identify the local summits and bottoms for computer programs. However, the number of summits counted using this orthodox method is strongly dependent on the sampling distance. Moreover, that is not always applicable to multi-directional surface textures, because the area shape and size by which local maximums/minimums are extracted cannot be determined in an analytical way. On the other hand, so-called contour mapping method can be utilized for asperity identification. This is also influenced by the height separation between two adjoining contours and its computer program may be composed of complicated routines. It seems to be difficult for the latter to identify the microscopic local maximums/minimums. A comparison of the functional performance between the two methods has not been made. The consistent optimal conditions under which both methods provide us with similar results are preferably needed from a practical point of view.