Characterization of general systematic form errors for circular features

Abstract

Characterizing systematic form errors in manufactured parts is essential to generate meaningful feedback information for production process improvement, as well as to develop a reliable conformance checking method. A new iterative measurement data analysis method is presented in this paper to characterize general and significant systematic form errors in parts with circular features, with two mathematical models based on combinations of Fourier components. The method starts from the ideal circular geometry and progresses by adding the most weighted Fourier component into the current model. This iterative process first constructs the dominant model containing the dominant Fourier components in the unknown systematic form error through the partial F-ratio test. The process then goes further to construct the significant model by capturing the remaining significant Fourier components through randomness tests that include the serial correlation and chi-square tests of the fitted residuals. It has been demonstrated from the analysis of extensive simulated and experimental data that the present method is able to characterize a wide variety of systematic form errors. The effect of random form error on the reliability of the systematic form error characterization has been examined and found to depend on its relative magnitude with the systematic form error.