Effect of pore characteristics and chloride binding on time-dependent chloride diffusion into cementitious material

© 2016 Australasian Corrosion Association. All rights reserved. Chloride ingress into concrete can be modelled using semi-empirical estimations of chloride diffusivity, D, from the power law D=D0(t/t0 )-m, where t is the time, t0 is the reference time when D=D0 and m is a fitting parameter. Incomplete knowledge of the reasons associated with the fitting parameter (m) could generate penetration profiles that become meaningless for the prediction of infrastructure service life. In this study, the effect of pore closure (physical effect) and chloride binding (chemical effect) on the chloride diffusion was investigated. Ordinary Portland cement (OPC), meeting the requirement of Australian General Purpose (GP) cement, was used. Both a through-diffusion test and an "in-and-out" diffusion test were conducted to monitor the time-dependent chloride diffusion through cement paste cured for different times. The through-diffusion test revealed overall chloride diffusion behaviour at different stages of cement hydration, which was considered to be a combination of both physical processes and chemical retention processes during the diffusion. The "in-and-out" test differentiated better the contribution of the physical and chemical processes on the chloride diffusion at different stages of cement hydration. Test results show that the reduction of chloride diffusivity is greatest at early ages of curing and may be attributed to the significant reduction of porosity and establishment of discontinuities between capillary pores. The results suggest that at very early curing times, capillary transport dominates, but then D follows the power law as described above. It was also observed that chloride binding capacity of the OPC increased with increased curing time from 1 to 90 days.