Spatial variability of concrete electrical resistivity and corrosion rate in laboratory conditions

Abstract

Spatial variability of reinforced concrete structures is responsible for part of the uncertainties that are associated with most parameters in structural durability and reliability models. Consequently, its effects should not be overlooked on structural reliability analysis for deteriorating structures. The present study characterized the spatial variability of concrete electrical resistivity and corrosion rate by means of descriptive statistics, estimations of probability density functions and correlation matrices. Data for these two corrosion parameters was obtained by monitoring two 50cm×40cm×10cm reinforced concrete slabs on a laboratory environment. Chloride was added to the mix of one of these specimens in order to promote corrosion. The main objective of the test set was to contribute in quantifying and characterizing the spatial variability of these two corrosion parameters in order to improve inspection planning of concrete structures and also probabilistic models that employ these parameters as variables. The variation obtained for corrosion rate and resistivity was considerable taking into account the small size of specimens tested. When compared with concrete resistivity, corrosion rates presented a larger spatial variability, especially when there is active corrosion. Data distribution for resistivity was found to be positively skewed. Correlation between resistivity values obtained in different locations of the slabs was strong, decreasing with distance as expected. As such, the spatial autocorrelation was determined for the concrete resistivity. On the contrary, adjacent measurements of corrosion rate showed no correlation between themselves. A better knowledge of the variables behaviour is fundamental for the quality of any probability based model.