An ANOVA method of evaluating the specification uncertainty in roughness measurement

The specifications of roughness used in industry are normally incomplete, and the incompleteness can induce a significant uncertainty, called specification uncertainty. It’s important to know the magnitude and effect of this uncertainty, but there are yet no standard methods of evaluating the specification uncertainty. In this paper, we propose an ANOVA method to estimate the specification uncertainty. In this method, ANOVA is used to separate specification uncertainty from measurement uncertainty, and the sampling method of GR&R (gauge repeatability and reproducibility) is applied. A case study is given to demonstrate how to use this method to evaluate the specification uncertainty of measuring roughness with PGI (Phase Grating Interferometer) when the filter type is not specified.