Evaluating visual and tactile perceived quality of machined surfaces using multiscale slope and curvature analyses

Abstract After finish milling in die and mould manufacturing, the method used for quality control is usually visual and tactile inspection of the surface finish by the operator. The purpose of the inspection is to link the manufacturing conditions to the appearance of the surface. The perceived quality of a surface is related to how light is reflected by that surface which in turn depends on geometrical features, such as distribution of heights (i.e. surface roughness) as well as slopes and curvatures. Prior work relating perceived quality to surface properties has almost exclusively focused on roughness. The present study re-examines data from a previous study with the objective to investigate how the geometric topography characterisations multiscale curvature and slope analysis could be useful for relating surface texture both to how finish milled metallic surfaces were processed as well as how the appearance was perceived. It was found that large variation in surface curvatures and slopes, as well as sharp curvatures and steep slopes, generate lower perceived quality, and that this effect was most prominent in a certain range of scales. Here, it was suggested to use the slope parameter Total variation as a link between manufacture and function since it has good discriminatory power regarding tool and workpiece material used as well as discriminatory power and very good statistical correlation (Spearman rank coefficient ρ = -1) regarding perceived appearance when calculated at a scale around 20 μm.