Adrian-Alexandru Pîrvan
The calcium silicate hydrates (C-S-H) are without doubt one of the most important hydration products in a hardened cement paste. Giving the complexity of the microstructure that forms by hydration of ordinary Portland cement (OPC) and the more recently used cements with supplementary cementitious materials (SCMs), studying the main feature and properties of C-S-H is not an easy task. Various methods for precipitating synthetic C-S-H with controlled composition have been developed, but the resulting products lack a solid microstructure and the characteristic pore network found in cementitious materials.The research plan of this thesis includes using a reactive dicalcium silicate binder which possess two advantages, compared to alite or OPC: it hydrates fully is less than 7 days and very low amounts of portlandite form, especially at lower water-to-binder ratios. The evolution of the C-S-H composition with time and w/b in the hydrated C2S microstructure was investigated. A slight drop in the Ca/Si ratio of C-S-H with hydration age was observed. A more significant decrease occurs when the w/b was increased from 0.35 to 0.80, and a higher amount of CH was precipitated in the hydrated matrix. With increased content of alkali introduced in the mixing water, the Ca/Si ratio of the resulted C-S-H further decreased from almost 2.0 to 1.80 and more portlandite formed. This was correlated with the metastable character of the high Ca/Si ratio C-S-H with respect to the pore solution and supersaturation with respect to C-S-H and CH are expected to play an important role.The effect of two frequently used SCMs, micro-silica and metakaolin, on the C-S-H microstructure was investigated by hydrating mixtures of the reactive C2S with different amounts of the two SCMs. The Ca/Si ratio of C-S-H decreased with higher amount of silica/metakaolin used and at 20% replacement level, the Ca/Si ratio at 28 days was close to 1.60. With hydration age up to 9 months, more silica and metakaolin is consumed, the specific surface area of C-S-H is reduced and the porosity is refined. A pozzolanic reaction was thus identified in a system without portlandite available to be consumed in the reaction.Transport properties through hydrated matrix are particularly interesting, as chloride penetration is related to degradation of steel reinforced concrete elements. The C-S-H microstructures were tested for both chloride binding and chloride migration. A comparison with hydrated white cement microstructures and synthetic C-S-H powders was presented. In terms of chloride binding, the hydrated C2S adsorbed less than the synthetic C-S-H, but some systems retained slightly more than the white cement paste. The amount of adsorbed chloride was significantly reduced in samples containing silica and metakaolin. Concerning the chloride migration, effective diffusion coefficients for all C2S and white cement systems were calculated. Migration of chloride through the plain C2S microstructure was found to be much faster than through a hydrated white cement at the same water-to-binder ratio, but the performance could be improved by reducing the water-to-binder ratio and through the use of micro-silica and metakaolin. These results are showing the strong potential of the reactive C2S to be used for the study of C-S-H phase with various compositions in isolation and for the investigation of various transport phenomena associated with concrete durability issues.
EPFL2022
