Influence of sample size on flatness estimation and uncertainty in three-dimensional measurement

Abstract

To respond correctly to its function in a mechanism, a mechanical part should have a good dimensional and geometric quality. In three-dimensional measurement, the preferred tool of control is currently the coordinate measuring machine CMM, the evaluation of flatness defect of a part is a very common task in this area, and the decisions made based on the measurement result may be outliers when the uncertainty of measurement result is not taken into account. The evaluation of uncertainty is a delicate task in this area given the diversity of parameters that relies in. The number of points collected on the surface is one of the influencing factors. The choice of this parameter is left to the operator, and the value taken by this parameter affects both the estimated flatness and the associated uncertainty. In this paper we present the effect of a number of points collected during the inspection of a surface on the estimated flatness and the associated uncertainty. A new identification method is presented here. The flatness deviation uncertainty is quantified by applying the law of propagation of uncertainties. The treated data sets are the result of experimental measurements performed on a CMM by measuring an industrial part.