This paper presents a 3-D surface topography model that has been developed to analyze and represent the spectrum of components of surface topography ranging over shape, waviness and roughness in a way suitable for generating macro- and micro-level automated surface finishing commands. Results of 3-D surface topography analysis such as comparison of desired and actual surface topography must be combinable with surface finishing processes in ways that permit the generation and output of commands, thereby causing the actual surface topography to converge towards a desired surface topography.An efficient filter, 3-D motif filter, with optimal cut-off length has been developed for roughness elements first, and then extended to waviness elements.A neutral surface shape method approximation for the machined surface shape has been proposed. Surface shape error is then eliminated as the deviation of the approximated neutral surface shape and the design surface shape.After the motif filtering, a B-spline fitting is used to represent the surface topography data within each 3-D motif cell after surface shape regeneration. The data reduction technique in 3-D motif filter methods has been shown significantly by reducing a great amount of inspection data to several elements and each element is suitable for use in surface finishing control. Some limitations of the 3-D motif filter method are discussed by the selection of different machined lay direction. The use of the developed 3-D surface topography model on automated surface finishing processes was performed on a platform constituting a 3-axis CNC machining center, inspection probe and a electrical grinder installed on the spindle housing of CNC machining center. Experimental results from an automated surface finishing system (ASFS) with in-process surface topography acquisition for mold and die finishing have shown the optimistic use of the developed surface topography model for industrial applications.
International Journal of Machine Tools & Manufacture 38(5-6), pp.543-550