Similarities and probability distributions of chloride convection zone depth in concrete exposed to cyclic drying-wetting environments

Abstract

A methodology using parabolic fitting was proposed to regress three points in chloride profile and determine the depth of convection zone in concrete. Based on exposure tests, the influences of environmental exposure, exposure time and mixture proportion of concrete on probability distributions and statistical parameters of convection depth were studied. The similarity in probability distribution of convection depth was analyzed by the Kullback-Leibler divergence. It was found that the mean, coefficient of variation and distribution type of convection depth are significantly influenced by environmental condition. Convection depth in natural environment mostly follows the Gumbel (Type I) distribution, while it follows the Beta distribution in simulated environment. The probability density functions of convection depth in simulated environment at early exposure have high similarities with those in natural environment at later exposure.