Methodology for the assessment of measuring uncertainties of articulated arm coordinate measuring machines

Articulated Arm Coordinate Measuring Machines (AACMMs) have gradually evolved and are increasingly used in mechanical industry. At present, measurement uncertainties relating to the use of these devices are not yet well quantified. The work carried out consists of determining the measurement uncertainties of a mechanical part by an AACMM. The studies aiming to develop a model of measurement uncertainty are based on the Monte Carlo method developed in Supplement 1 of the Guide to Expression of Uncertainty in Measurement [1] but also identifying and characterizing the main sources of uncertainty. A multi-level Monte Carlo approach principle has been developed which allows for characterizing the possible evolution of the AACMM during the measurement and quantifying in a second level the uncertainty on the considered measurand. The first Monte Carlo level is the most complex and is thus divided into three sub-levels, namely characterization on the positioning error of a point, estimation of calibration errors and evaluation of fluctuations of the `localization point'. The global method is thus presented and results of the first sub-level are particularly developed. The main sources of uncertainty, including AACMM deformations, are exposed.