A tolerance interval based criterion for optimizing discrete point sampling strategies

In order to compute geometric tolerances, the distance between the actual geometry and the nominal one has to be computed, and this computation requires the selection of a sampling strategy. Sampling strategy consists in deciding the number and the locations of the points that have to be measured on the actual surface. This paper presents a new approach for selecting the optimal locations of the points for any given sample size. The proposed procedure is based on estimating the “manufacturing signature”, that is the systematic pattern left by the manufacturing process on the machined items. In particular, the proposed approach is based on using regression for estimating the manufacturing signature and then selecting the measurement point locations by minimizing the distance between the maximum and the minimum points of the regression-based tolerance interval.Performance of the proposed procedure are compared with those obtainable by selecting different strategies with reference to a case study related with roundness data. Given that the approach proposed outperforms the competitor ones, a robustness study is finally presented to show that the method presented is robust to violation of the assumption behind linear regression models.