Transfer of measurement uncertainties to reduce the effort required for proof of suitability (MessAgE)

点击翻译按钮获取中文摘要

Each measurement is inherently uncertain. If uncertain measurement data are further used, derived statements are also subject to uncertainty and can lead to erroneous decisions. If a produced characteristic is close to the specification limits and if the measurement uncertainty is too high during inspection, it cannot be determined with certainty whether the characteristic lies within or outside the specification. This can lead to parts lying within the specification and being sorted out during inspection (α error) or vice versa (β error). Since the risk of wrong decisions cannot be determined and thus cannot be controlled without knowledge of the measurement uncertainty during measurement data collection, measurements without measurement uncertainty specifications are worthless. In order to prevent economically harmful wrong decisions, numerous standards and guidelines exist in the production environment, which require the determination of measurement uncertainty and set limits for the suitability of measurement and inspection processes. Due to the networking of digital products and resources (including machining centers, assembly systems) and the associated focus on cyberphysical systems, the number of sensors and measurement processes used is constantly increasing. With the increasing number and complexity of inspection processes, the effort for corresponding proofs of suitability is growing regardless of the industry. In the research project applied for, a procedure is being developed which reduces the effort required for determining measurement uncertainty and for providing proof of suitability. For this purpose, methods are first developed which quantify the mathematical/geometric, technological and functional similarity of measurement processes. Based on the similarity assessment, a method is developed to transfer measurement uncertainties from one measurement process to other measurement processes. By considering the resulting transfer error, the transfer of the measurement uncertainty is secured. The result is a generic similarity definition for measurement processes and a procedure for transferring known uncertainties of existing measurement processes to similar measurement processes in order to reduce the effort for determining the measurement uncertainty.