A Factor Analysis Approach for Robust Inspection of Circular Features with Lobing Errors

摘要

Circular and cylindrical features typically have more than one GD & T parameter (form, position, orientation, and size) that needs to be estimated simultaneously using a single inspection procedure. In such instances, inspection methods have to be robust, so that they have good overall performance in simultaneously assessing all the GD & T parameters of interest, and an optimal sample size is desirable. This paper proposes a method to deduce the optimal sample for robust inspection of form, position, and size of circular features with systematic lobing errors. Equispaced sampling, employing a least-squares fitting, has been assumed for inspection. A novel Factor Analysis approach has been used to study the correlation tructure of the performance measures (bias and standard deviation), and subsequently, the robust sample size was deduced. Results obtained are in agreement with conventional recommendations for circularity inspection. Directions for future research have been outlined.