Propriétés d’usage et mécanismes d’hydratation du système ternaire Ciment Alumineux – Sulfate de Calcium – Laitier de Haut Fourneau à haute teneur en sulfate de calcium : De l’approche expérimentale à la modélisation

摘要

The potential of activation of a Ground Granulated Blast Furnace Slag has been evaluated into a ternary system comprising of a Calcium Sulfate as major component and a Calcium Aluminate Cement. This system is not well known and the main goal of this study is to determine its main advantages and limitations. From the usage properties point of view, fast setting and initial strengths are governed by the ettringitic binder. Then, and only if the dehydration is avoided, slag reacts. In this case, slag contributes to the increase of mechanical performances at medium and long terms and to limit the expansion under water. From an understanding point of view, the original experimental approach reveals the potential of the slag. It is based on a comparison of performances with different types of curing methods. The potential of hydration of the slag is amplified when the system is dehydrated during a few days and then rehydrated. However, slag contribution is complex to establish because analytical methods to follow-up slag consumption (such as X-ray Diffraction and Differential Scanning Calorimetry analysis) are not well adapted. So, to understand the hydration mechanisms, indirect approaches are used. Semi-quantitative methods by X-ray Diffraction to follow-up the mayenite, calcium sulfate, syngenite and ettringite, ThermoGravimetric Analysis to measure degree of hydration and Mercury Intrusion Porosimetry to identify microstructural changes have been carried. The most important difficulty concerns the identification and quantification of amorphous phases such as slag, C-S-H and AH3. That is why a modelling approach is necessary to understand the role of each compound in the ternary mixture and in particular the obvious contribution of slag during rehydration test. This modelling approach increases knowledge of the physical and chemical phenomena in this ternary binder. It is useful to explain the observed macroscopic properties such as strength and helps to determine the kinetics of hydration in porous environment. Even if this model is still under development, it has allowed identifying the sequences of hydration (ettringitic binder reacts at very initial time, then anhydrite transforms into gypsum and slag reacts at long term) and confirms therefore that the reaction of slag is slow.