Consideration and Impact Assessment of Measurement Uncertainty in the Context of Tolerance Analysis

Abstract

Abstract Virtual product validation ensures that products fulfil their function under different varying conditions. Within the framework of virtual product validation, variation simulation represents the specific area that deals with the consequences of geometric part variations on functional key characteristics. In addition to the nominal geometry of the parts, the permissible part deviations in the form of tolerances, the joining sequence and process variations during assembly are required to simulate the effects of geometrical part deviations on assembly or product key characteristics. If the required quality targets are not achieved, the part tolerances usually have to be tightened, which goes hand in hand with increased production costs. The achievable part tolerances depend, among other things, on the material used, the workpiece dimensions, the manufacturing process, but also on the interaction between these factors. Therefore, the prediction of specific manufacturing process dependent deviations is hardly possible. However, if predictions should be made as exactly as possible, there is also the possibility to integrate measurement data directly or indirectly into the tolerance analysis. This has the advantage that material- and production-specific deviations can be considered in the best possible way for a certain part geometry. However, if measurement data is integrated into tolerance analysis, the problem arises that, in addition to the actual component deviation, the measurement uncertainty as part of the measurement result, must be implicitly determined in order to also be taken into account. Conversely, the tolerance analysis results are also influenced by the measurement uncertainty. To tackle this issue a novel procedure is presented which allows the quantification of the influence of the measurement uncertainty on the result of the tolerance analysis. In addition, it is shown how the measurement uncertainty is determined, whereby in particular the single point measurement uncertainty is dealt with. Since the measurement uncertainty can be different for each measuring point, Skin Model Shapes are used for tolerance analysis in order to have the possibility for defining point specific information. The developed procedure is then applied to a suitable case-study. CT measurements are used as the measurement method for determination of the single point measurement uncertainties. Finally, different scenarios for the tolerance analysis are compared in order to quantify the influence of the measurement uncertainty in the best possible way.