Structure and Mechanical Properties of Cement and Intermetallic Compounds via ab-initio Simulations

Abstract

Calcium silicate hydrates comprise a class of minerals formed synthetically during Portland cement hydration or naturally through various geological processes. The importance of these minerals is immense since they are the primary binding phases for Portland cement derived construction materials. Efforts spanning centuries have been devoted to understand the structural aspects of cohesion in these minerals. In recent years, the focus has progressively turned to atomic level comprehension. Structurally these minerals can range from crystalline to highly disordered amorphous phases. This thesis focuses upon unraveling the nature of chemical bonding in a large subset of calcium silicate hydrate (CSH) crystals. Thus their electronic structure was calculated and bonding mechanisms were investigated quantitatively. Results highlight a wide range of contributions from each type of bonding (Si-O, Ca-O, O-H and hydrogen bond) with respect to silicate polymerization, crystal symmetry, water and OH content. Consequently, total bond order density (TBOD) was designated as the overall single criterion for characterizing crystal cohesion. The TBOD categorization indicates that a rarely known orthorhombic phase Suolunite is closest to the ideal composition and structure of cement. Present work finds the relationship of partial bond order density (PBOD) of each bond species, especially HBs to the mechanical properties of CSH crystals. This can be used as a basis to validate existing C-S-H models and to build improved ones. This work goes further and validates the recently proposed models (2014) for C-S-H (I) phase on the same basis of proposed electronic structure parameters. Then the respective Calcium aluminosilicate hydrates C-A-S-H (I) phase models are proposed. Finally, these results lead to improved interpretations and construction of realistic atomistic models of cement hydrates. Ab initio molecular dynamics (AIMD) could be vital to solve critical problems in complex structural material such as cement. However, it is far too early to be applicable for cement. Thus, this study used intermetallic compounds as a test case to develop new AIMD methods. In light of this objective, a direct method to calculate high temperature mechanical properties was devised for Mo₅Si₃ (T1 phase) and Mo₅B2Si₃ (T2 phase). It was found that thermal expansion anisotropy (TEA) of T1 phase is captured by this simulation. It was also found an AIMD method to reduce TEA of Mo₅Si₃ (T1 phase) by strategic alloying. With further research these methods may be transferrable to cement and may allow optimizing the performance of hydraulic cements.