Abstract
Abstract. During the manufacturing of work pieces, geometrical deviations from the intended nominal geometry of the designer are inevitable. The procedure of conformance testing defined in ISO 14253-1:2018-07 is used to ensure the function of a work piece by verifying the geometrical compliance with pre-defined tolerance specifications. Depending on the measurement setup used for the validation step, it is possible that the local measurement uncertainty is too large in order to provide a meaningful conformance evaluation. This paper aims to demonstrate the complete workflow of the determination of the locally defined single point uncertainty and its components (systematic and random measurement error, respectively) for a given measurement task. It was shown for an optical measurement setup in combination with an industrial X-ray computed tomography reference measurement system that different necessary colouring methods of polymer (POM) gear wheels, which are required to enable measurements using structured-light scanning, have a measurable influence on the local distribution of the measurement uncertainty. Because of the fact that the presented method is dependent on a discrete surface sampling, the effects of different polygonization settings during the creation of the areal measurement result were evaluated in order to rate the reduced data complexity against the hereby possibly increased measurement uncertainty. The gained information regarding the local measurement uncertainty of a measurement setup can then be used for downstream processes in various use cases, e.g. for the improvement of holistic tolerance simulation models or the improvement of geometrical measurements using weighted regression analysis. Additionally, the visualization of the areal distribution of the measurement uncertainty enables a powerful tool to optimize the used measurement setup.
Highlights
The results of any geometrical measurement of a work piece and the subsequent comparison against the nominal geometry (CAD – computer-aided design – model) as defined by the designer can be divided into three different contributions: (a) the geometric deviations of the measurement object with respect to the nominal geometry, (b) the systematic measurement error of the measurement system and (c) the random measurement error of the measurement system
The goal of this paper is to present the framework of the single point uncertainty as a useful tool for the quantitative evaluation of a specific measurement setup using the example of an optical polymer gear measurement task
It was shown that the optical measurement of POM gear wheels using structured-light scanning can in principle be realized by colouration during the manufacturing process
Summary
The results of any geometrical measurement of a work piece and the subsequent comparison against the nominal geometry (CAD – computer-aided design – model) as defined by the designer can be divided into three different contributions: (a) the geometric deviations of the measurement object with respect to the nominal geometry, (b) the (signed) systematic measurement error of the measurement system and (c) the (unsigned) random measurement error of the measurement system. The golden rule of metrology states that the measurement uncertainty shall be less than 10 % to 20 % of the tolerance (Knapp, 2001; Berndt et al, 1968). This requires a close observation of the achieved measurement uncertainties for each geometric verification task. – A single reference measurement of the measurement object (reference geometry) performed by a reference measurement system in the sense of the International Vocabulary of Metrology (VIM) (Brinkmann, 2012) which is capable of recording areal measurements of the chosen measurement object (see the note above)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
