Scrivener, KarenLothenbach, BarbaraChitvoranund, Natechanok2021-05-272021-05-272021-05-27202110.5075/epfl-thesis-8783https://infoscience.epfl.ch/handle/20.500.14299/178380Two low CO2 cement systems are investigated: one made with the use of supplementary cementitious materi-als (SCMs) (metakaolin and limestone) and the other an alternative binder, calcium sulfoaluminate cement (CSA). These two low CO2 cements contain a similar higher alumina content compared to Portland cement (PC), but they have con-trasting chemical hydration reactions. It leads to the changes in phase assemblage, microstructure and properties. There have been previous studies on these materials separately mainly at room temperature. But here we make a systematic comparison looking at several temperatures and relating the phase assemblages and microstructure to strength and transport properties. Temperature has a strong influence on the aluminate phases such as ettringite, monosulfate and strätlingite. Thermody-namic modelling with the current CEMDATA database was used over a wide range of realistic curing temperatures. It is a good tool to predict phase assemblages in the equilibrium state, especially for qualitative predictions. The accuracy of the characterization methods is an important factor to obtained the best inputs. Accurate measurement of the degree of reaction of amorphous phases such as metakaolin is challenging. Another issue is to obtain a realistic composition of the C-A-S-H phase. A more homogeneous microstructure was observed in the metakaolin blends compared to the PC reference and the CSA system. It improves the transport properties. However, the higher alumina content of the blended systems seems to inhibit the hydration of belite, resulting in a lower the compressive strength, especially for the CSA system at early age. The correlations between the hydration, microstructure and properties were explored with various techniques. For the Portland based systems, the combined water fraction correlates well with the compressive strength despite some devia-tion in the blended systems. However, this concept does not appear to relate well to strength for the CSA system. The mini-migration migration method can be used to determine the resistance of chloride, especially in the PC systems. A simple method of effective conductivity can be applied as a rough indicator. The transport properties in the CSA system are more difficult to measure with accelerated methods.enmetakaolinlimestonecalcium sulfoaluminate cementtemperaturephase assemblagethermodynamic modellinghydrationmicrostructurecompressive strengthchloride ingress.thesis::doctoral thesis