Determination of uncertainty of measurement for penetration testing

Abstract

Paper presented at the 30th Annual Southern African Transport Conference 11-14 July 2011 "Africa on the Move", CSIR International Convention Centre, Pretoria, South Africa.

Any measurement has an error associated with it. The parameter that quantifies the boundaries of the error of that measurement is known as the uncertainty of measurement. Without determining the uncertainty of measurement, any measurement lacks merit. An assessment of uncertainty of measurement is carried out for the standard test method for penetration of bituminous materials as prescribed by ASTM D5. The assessment of uncertainty of measurement is performed using the principles stated in the IS0 Guide to the Expression of Uncertainty in Measurement (IS0 GUM). A model was formulated for the purposes of determining the uncertainty of measurement, incorporating the equipment (based in the advanced testing laboratories at the CSIR), environment and test procedure. Sources of measurement variation were identified and isolated within the model. Furthermore, these sources were classified as either random or systematic, as well as whether they were subject to type A or type B analyses. Experimental processes quantified the various sources of uncertainty and an overall uncertainty of measurement was calculated for the test method. This paper illustrates how IS0 GUM may be applied to determine the accuracy of the penetration test method. The results can be used to refine the penetration test procedure and tighten up on equipment specifications in order to reduce the level of test uncertainty. In addition, this paper may serve as an example of how to extend IS0 GUM to other laboratory test methods.