Validation of the sensitivity analysis method of coordinate measurement uncertainty evaluation

Abstract

The paper presents the results of the tests carried out to validate a new method for evaluating the uncertainty of coordinate measurements categorised as the Sensitivity Analysis (SA). This method concerns measuring dimensions and geometrical deviations. Measurement uncertainty is evaluated on the basis of information given in the Maximum Permissible Error (MPE) formula for a Coordinate Measuring Machine (CMM). Measurement models express the measured characteristics as a function of differences of coordinates of a small number of appropriately selected points of the workpiece. If reverification test results for the CMM used are available, then the estimated uncertainty takes into account the actual accuracy of the CMM. General formulae are given to calculate the uncertainty of measurement of a circle diameter and coaxiality. The relevant experiment is based on ISO 15530-3 recommendations. Acalibrated cylindrical square was used for validation. 17 circles’ diameters and 84 different combinations of datum length and distance of the toleranced element from the datum for measuring coaxiality were adopted as validated characteristics. The validation results are presented in tables and graphs and the chi-square test for equality of variances was used to confirm that the method is correct. The validation results are positive.