Go back to Part I Chapter 7 Sec. (3b) Go back to Table of Contents
Experimental and computer modeling studies are applied to determining the influence of silica fume on the microstructure and diffusivity of cement paste. It is suggested that silica fume modifies the inherent nanostructure of the calcium silicate hydrate gel (C-S-H), reducing its porosity and thus increasing its resistance to diffusion of both tritiated water and chloride ions. Because the pores in the C-S-H are extremely fine, the relative reduction in diffusion depends on the specific diffusing species. Based on the NIST cement hydration and microstructural model, for tritiated water diffusion, the reduction in the diffusivity of the gel caused by silica fume is about a factor of five. For chloride ions, when a diffusivity value 25 times lower than that used for conventional high Ca/Si ratio C-S-H is assigned to the pozzolanic lower Ca/Si ratio C-S-H, excellent agreement is obtained between experimental chloride ion diffusivity data and results generated based on the NIST model, for silica fume additions ranging from 0 % to 10 %. For higher addition rates, the experimentally observed reduction in diffusivity is significantly greater than that predicted from the computer models, suggesting that at these very high dosages, the nanostructure of the pozzolanic C-S-H may be even further modified. Based on the hydration model, a percolation-based explanation of the influence of silica fume on diffusivity is proposed and a set of equations relating diffusivity to capillary porosity and silica fume addition rate is developed. A 10 % addition of silica fume may result in a factor of fifteen or more reduction in chloride ion diffusion and could potentially lead to a substantial increase in the service life of steel-reinforced concrete exposed to a severe environment.
Keywords: Cement paste(D), diffusion(C), hydration(A), modeling(E), silica fume(D).