Evaluation of chloride penetration in high performance concrete using neural network algorithm and micro pore structure

Chloride attack is one of the major causes of deterioration of reinforced concrete structures. In order to evaluate the chloride behavior in concrete, a reasonable prediction for the diffusion coefficient of chloride ion, which governs mechanism of chloride diffusion inside concrete, is basically required. However, it is difficult to obtain chloride diffusion coefficients from experiments due to time and cost limitations. In this study, a numerical technique for chloride diffusion in high performance concrete (HPC) using a neural network algorithm is proposed. In order to collect comparative data on diffusion coefficients in concrete with various mineral admixtures such as ground granulated blast-furnace slag (GGBFS), fly ash (FA), and silica fume (SF), a series of electrically driven chloride penetration tests was performed. Seven material components in various mix designs and duration time are selected as neurons in a back-propagation algorithm, and associated learning of the neural network is carried out. An evaluation technique for chloride behavior in HPC using the obtained diffusion coefficients from the neural network algorithm is developed based on, so-called, Multi-Component Hydration Heat Model (MCHHM) and Micro Pore Structure Formation Model (MPSFM). The applicability of the developed technique is verified by comparing the analytical simulation results and the experimental results obtained in this study. Furthermore, this proposed technique using the neural network algorithm and micro modeling is applied to available experimental data for verification of its applicability.