An Innovative Conformity Criterion for Assessment of the Concrete Strength Under Uncertainty Conditions

Abstract

One of the most important sections of a new version of EN 206 is the concrete strength conformity control. In spite of some changes in the assessment procedure introduced in the new version of this standard (A, B, C conformity assessment methods), the main problem of this document is related to uncertainties in conformity assessment of the concrete strength based on a limited number of test results (less than 15) recorded at the initial production stage. As it was shown in numerous international publications, conformity criteria for initial production are often vague and based on arbitrary suggestions, sometimes without coherent probabilistic background. These criteria are strongly influenced by type of probability density function in population, standard deviation, shape of specimen.

The paper presents an innovative criterion for conformity assessment of concrete strength which was developed based on non-parametric order statistics with some modification of the estimation procedure. Based on non-parametric confidence interval, estimator for p = 0.05 quantile was obtained. Probabilistic transformation (Laplace inverse function) and subsequent three-point linear approximation has been applied to the proposed non-linear model for the quantile estimator. Proposed criterion was positively verified based on numerical analysis (an AOQL concept with Monte Carlo simulations) as well as on data of concrete strength records of the real production.