| dc.description.abstract | Cement production stands for approx. 8% of the man-made CO2 emissions. One of the most common ways to reduce CO2 emissions in cement production is by replacing part of the clinker with supplementary cementitious materials (SCMs) to produce Portland composite cements.
Todays’ Portland composite cement produced at the Norcem cement plant in Brevik is a CEM II/B-M containing 18wt.% fly ash (FA) and 4wt.% limestone (L). However, the availability of FA is decreasing in Europe leading to the unavoidable need for other SCMs with properties similar to FA.
This PhD thesis investigated the impact of a selection of SCMs and curing temperatures on the phase assemblage and porosity of hydrated composite cement pastes and how that in turn affect the compressive strength of the respective mortar samples.
Cement pastes with a w/b-ratio of 0.50 and mortar samples with a water-cement-sand ratio of 0.50:1:3 were prepared and cured sealed at 5, 20, and 38 °C for at least 180 days to ensure a sufficient reaction of the SCM. Composite cements with a composition of 78wt.% Portland cement (PC), 18wt.% supplementary cementitious material (SCM), and 4wt.% limestone (L), as well as two reference cements, one containing 78wt.% PC and 22wt.% limestone (PC-L), and the other being a pure PC were investigated. Three SCMs were used in this study, i.e., a siliceous fly ash (FA), a silicomanganese (SiMn) slag (S), and volcanic pozzolan (VP).
The phase assemblage was investigated in-depth by determining the C(-A)-S-H composition, the degree of reaction of the components, the mass of hydration phases, and the porosity characteristics. The compressive strength was determined on the mortar samples. The investigations applied a multi-technique approach including measuring techniques, i.e., XRD, TGA, SEM-EDS, water suction, MIP, DVS, and compressive strength testing, and thermodynamic modelling using GEMS and mass balance calculations.
The study of the C(-A)-S-H composition for the SCM-containing cements showed that the C(-A)-S-H has a lower Ca/Si and S/Si, but higher Al/Si ratio than the C(-A)-S-H of the PC as expected, because the SCMs contribute with silicates and aluminates. With increased curing temperatures, the SCMs show an enhanced reaction and a further decrease in the Ca/Si and S/Si ratios, and a further increase in the Al/Si ratios. The quantification of the changes in the C(-A)-S-H composition were crucial, as it had a considerable impact of the mass balance calcualtions of these systems.
The degree of reaction of the SCMs was determined both using SEM combined with image analysis and using mass balance calculations combined with portlandite quantification using TGA. Similar reaction degrees were obtained using the two different methods. In general, similar reaction degrees weredetermined for the three SCMs despite their different origins and chemical compositions. The reaction degrees were enhanced with increased curing temperatures, i.e., 20, 36, and 47% for FA, 34, 43, and 60% for S, and 35, 60, and 70% for VP when cured at 5, 20, and 38 °C, respectively. The degree of reaction of the SCMs are included in the mass balance, which enabled the calculation of the mass of the hydration phases formed.
For the PC and PC-L reference mortars, the compressive strength gradually decreased with increasing curing temperature. However, a different temperature dependency of the compressive strength was measured for the SCM-containing mortar samples, and the highest compressive strength was measured after curing at 20 °C, and lower strengths were observed for both 5 and 38 °C. The deviating temperature dependencies of the compressive strength between the reference and the SCM-containing mortars was investigated by relating the compressive strength to the degree of reaction of the components, the amount of C(-A)-S-H, and the total volume of the solids (unreacted solids and hydration phases) both with and without the fine porosity associated with C(-A)-S-H. The temperature dependency of the compressive strength was reflected in all these parameters for the PC and PC-L references. However, in the case of the SCM-containing composite cements, none of these characteristics could explain the curing temperature dependency of the compressive strength. One of the main differences between the reference (PC-L) and the SCM-containing cements was an enhanced reaction degree at higher curing temperatures.
Correlating the porosity characteristics determined using a range of porosity techniques to compressive strength showed a clear link for the PC and PC-L references. However, this was not observed for the SCM-containing cements. There seemed to be a need to differentiate the pore size range when investigating the porosity characteristics of SCM-containing cements because the gel porosity may have a lower impact on the compressive strength than capillary and macro porosity. The differentiation of the pore sizes enabled to show the contribution of the SCMs to the compressive strength, but did not explain the temperature dependency of the composite cements.
In a last step, the impact of the curing temperatures and SCMs on the C(-A)-S-H density was calculated. Neither for the PC and PC-L references nor for the SCM-containing cements was a clear temperature impact on the C(-A)-S-H density observed. This might be due to the limited range of curing temperatures used. In the presence of the SCMs, the C(-A)-S-H densities were slightly higher than in PC-L meaning that the pozzolanic reaction of the SCMs led to a densification of the C(-A)-S-H. Though, the C(-A)-S-H densities in the composite cements were considerable lower compared to the PC reference indicating that the space available for the C(-A)-S-H to form, e.g. effected by the w/b-ratio, had a considerable impact on the C(-A)-S-H density.
The variations in the hydrate assemblage and the porosity did not explain the curing temperature dependency of the compressive strength of SCM-containing composite cements, as they did for the PC and PC-L references. Further research on the dependency of the C(-A)-S-H intrinsic strength on its composition and microstructure is needed to understand how the curing temperature impacts the compressive strength of SCM-containing composite cements.
The present thesis showed that the composite cements containing 18wt.% of the novel SCMs, i.e., iMn slag and volcanic pozzolan, had a similar response to variations in the curing temperature and a similar impact on the hydrate assemblage and porosity as the conventional composite cements containing fly ash. Thus, from the perspective of the hydration at later age, the composite cements containing the novel SCMs are valid alternatives for the currently used composite cements containing fly ash. | en_US |
