Influence of spatial correlation of core strength measurements on the assessment of in situ concrete strength

Many countries are experiencing an increasing need of checking the safety of existing structures. The assessment of structural capacity of RC structures strictly depends on the in situ compressive strength of concrete. The evaluation of this property is typically carried out by means of destructive tests on concrete cores taken from the structure. The experimental data is then interpreted using a relevant code to obtain a design strength value according to the required percentile and confidence. In this paper the principal international standards that deal with the statistical interpretation of data from concrete core tests are presented. Since it is reasonable to assume that concrete strength is a realization of a random field, the assumption of statistical independence of core test data is questioned. An extension of the classical theory of tolerance limits in the case of normally distributed correlated samples is thus proposed. Finally, application examples of this methodology are provided to illustrate some important implications of the spatial correlation of core test values on concrete strength estimations.