Non-linear dynamic compensation and re-sampling for tactile scanning measurement of curved surface topography based on GPS standards

Abstract Tactile scanning method is widely used for curved surface topography measurement. Owing to the non-linear transformation process from the vertical translational displacement of the measured point (the coordinates of stylus tip) to the angular displacement of the stylus tip (measured by the cylindrical diffraction grating), the non-linear error exists within the horizontal and vertical coordinates of the obtained sampling data. And the non-linear error normally rises with the increasing of the measurement range. To solve the non-linear problem, a tactile scanning measurement system based on cylindrical grating interference principle is constructed for curved surface topography; the originations of the non-linear error in this system are analyzed, the error compensation model is established, through which the non-linear error of the collected data is dynamically compensated. Subsequently, for the non-uniform distribution of the sampling points, data analysis method is investigated, in which non-uniform interpolation and re-sampling techniques are investigated to obtain accurate data with uniform interval, so that the analysis and evaluation of surface profile can be conducted based on GPS (Geometrical Product Specifications) standards. Experimental results prove the improvements.