Bisquare robust polynomial fitting method for dimple distortion minimization in surface quality analysis

In this paper, the bisquare robust polynomial fitting (BSRPF) method for surface texture features assessment was proposed. Plateau-honed cylinder liner bronze surfaces with additionally burnished microcavities (defined as dimple, holes, and/or cavities) were taken into consideration. Analyzed details were measured with stylus (Talyscan 150) or white light interferometer (Talysurf CCI Lite) methods. It was assumed that application of commonly used (in surface contact and noncontact measurement technology) polynomial fitting procedures did not allow to extract required features; distortion of microreservoir dimples for oil consumption assessment was falsely estimated when valleys were near edge or on the edge located. Application of bisquare robust polynomial fitting method caused a minimization of microflattens of dimples irrespective of valley distribution; the biggest degree of polynomial fitting method applied, the minimal distortions of oil reservoirs were noticed. The valley size (depth and width) and valley-to-valley distance were also taken into account.