DIFFUSION COEFFICIENT OF CHLORIDE IONS UNDER SIMULATED CONDITIONS

Abstract

Deterioration of concrete structures in Persian Gulf climate is a serious problem. Chloride ion penetration is responsible for such deterioration. Accelerated chloride penetration test (RCPT) and estimation of the diffusion coefficient is an effective way to predict the service life of concrete structures. In this research work, ordinary and silica fume concrete mixtures containing 0.35, 0.4, 0.45 and 0.5 water to cementitious materials ratios were used. Rapid chloride permeability test in accordance with ASTM C1202 and determination of diffusion coefficient of chloride ion penetration in accordance with bulk diffusion test under laboratory conditons simulated to Persian Gulf climate, were performed. Concentration of chloride ions in various depths of concrete specimens was measured using acid soluble chloride ASTM C114 test method. Results of this investigation show that silica fume reduces the chloride penetration and the diffusion coefficient in concrete mixtures. Three different models were made for RCPT results, diffusion coefficient, and correlation between these two parameters using a neural network system. Models are reliable and confirm the neural network ability for diffusion coefficient prediction. INTRODUCTION EXPERIMENTAL PROGRAMME TEST RESULTS AND DISCUSSION NEURAL NETWORK MODELS CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES