Similarity analysis of randomness in instantaneous chloride diffusion coefficient of concrete under different environmental exposures

This study addresses the temporal variability and randomness of concrete's chloride diffusion coefficient by proposing an analysis using stochastic measures for the instantaneous chloride diffusion coefficient (Dins) alongside Kullback-Leibler (K-L) divergence and Hamming distance for similarity assessment. Examination of diverse concrete specimens subjected to both natural tidal environments (NE) and simulated environments (SE) indicated that the environmental factors, primarily exposure setting and duration, substantially influence the diffusion coefficient's temporal behavior. SE with elevated temperature and salinity accelerated chloride penetration, leading to higher diffusion rate means than those in NE. Concrete in NE exhibited a log-normal distribution for Dins, while under SE, a normal distribution was observed. Variability was mainly driven by the exposure environment, with minor contributions from material components. Similarity analysis suggested that the stochastic character of Dins in SE after 120 days resembled those in NE at 600 days, as per K-L divergence, whereas Hamming distance showed the closest resemblance at 240 days in SE. The Material make-up showed negligible impact on distribution similarity, and the extending simulation exposure time had a limited effect on enhancing similarity of stochastic diffusion behavior across environments.