Cox, M G; Forbes, A B; Harris, P M; Peggs, G N (1998) Determining CMM behaviour from measurements of standard artefacts. NPL Report. CISE 15/98

Text cise15.pdf Download (11MB)

Abstract

This report is concerned with modelling and estimating the systematic error behaviour of co-ordinate measuring machines (CMMs). We describe two classes of model for these parametric errors, the kinematic models based on an axis-upon-axis build-up of behaviour and more general empirical models that can model any repeatable error behaviour. We then develop a comprehensive mathematical model for determining the parametric errors of a CMM from measurements of a standard artefact such as a baIlor hole plate and describe algorithms for determining relatively unbiased and efficient estimates of the parametric errors from such data. These algorithms have been implemented in software, engineered so that they can cater for any parametric error model, ball plate geometry and measurement strategy. The software also calculates standard uncertainties of the fitted parameters and related quantities. We use this software in numerical simulations to analyse the effectiveness of measurement strategies with respect to the statistical accuracy of parametric error estimates. The numerical experiments show that with a well-designed strategy it is possible to provide accurate estimates of the parametric errors using only a minimal amount of calibration information about the ball plate, but that other designs or estimators can give poor and/or biased estimates.
This work was supported by the National Measurement System Policy Unit of the Department of Trade and Industry, UK.

Item Type:	Report/Guide (NPL Report)
NPL Report No.:	CISE 15/98
Subjects:	Mathematics and Scientific Computing Mathematics and Scientific Computing > Modelling
Last Modified:	07 Sep 2018 14:26
URI:	http://eprintspublications.npl.co.uk/id/eprint/848

Actions (login required)

View Item