< back to main site


Reducing the effects of measurement noise when determining surface texture parameters.

Forbes, A B; Leach, R K (2009) Reducing the effects of measurement noise when determining surface texture parameters. In: 9th International Symposium on Measurement Technology and Intelligent Instruments (ISMTII-2009), 29 June - 2 July 2009, Saint-Petersburg, Russia.

Full text not available from this repository.


In this paper we describe methods that aim to remove or at least quantify the bias in determining surface roughness parameters due to the random effects associated with the measuring instrument. One approach is to use repeated measurements to allow the variation contribution of the instrument to be quantified and subsequently subtracted from the parameter estimates. For the case of Rq or Sq, these techniques are related to standard analysis of variance (ANOVA) approaches. One issue in using ANOVA in its classical formulation is that there is no guarantee that the assessed surface parameter will be greater than or equal to zero. Using a Bayesian formulation, the ANOVA approach can be extended to include prior information that will ensure that parameter estimates are physically meaningful. A key component of the formulation is to regard surface roughness parameters as scale parameters, necessarily positive, rather than location parameters, such as a position along an axis. Since the standard approaches for evaluating measurement uncertainty, including those described in the Guide to the Expression of Uncertainty in Measurement, are designed to be applied to location parameters, the assessment of uncertainty associated with surface roughness parameters has to be undertaken with care.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Subjects: Engineering Measurements
Engineering Measurements > Dimensional
Last Modified: 02 Feb 2018 13:15
URI: http://eprintspublications.npl.co.uk/id/eprint/4449

Actions (login required)

View Item View Item