Uncertainty of the variation in length of gauge blocks by mechanical comparison: a worked example.

This paper is a study on the determination of the measurement uncertainty for a relatively simple and widespread calibration task. The measurement of a special form deviation of gauge blocks using the so-called five-point technique is discussed in detail. It is shown that the mainstream treatment of the measurement uncertainty (i.e. propagation of uncertainties) cannot be applied to this problem for principal reasons; the use of supplement 1 of the GUM (Monte Carlo method) is mandatory. The proposed model equation is probably the simplest 'real world' example where the use of supplement 1 of the GUM not only gives better results, but gives results at all. The model is simple enough to serve as a didactical example. Explicit analytical expressions for the probability density functions, expectation values, uncertainties and coverage intervals are given which is helpful for the validation of dedicated software products. Eventually a complete avoidance of the standardized form parameters in calibration certificates is proposed. The statement of 'corner-deviations' would be much more useful, especially for the evaluation of key comparisons. [ABSTRACT FROM AUTHOR]