In recent years, cementitious materials and fibers have become widespread in concrete mixtures to enhance the mechanical and durability properties of engineering structures. However, incorporating such materials increases construction costs and requires extensive experimental trials to determine the optimal dosage. Consequently, researchers have conducted studies to predict cement-based materials’ mechanical properties using various techniques to avoid costly and labor-intensive experiments. Due to computer programming and hardware advancements, microstructure-based computational modelling techniques have garnered researchers’ attention. Hence, this study aims to propose a multi-stage model that predicts the mechanical properties of cement paste using unique input parameters, such as the chemical properties of cement and curing conditions, by integrating the hydration process and microstructure development. In the initial stage (Stage 0), the volume fraction of hydration products is predicted considering the formation of calcium silicate hydrates (C-S-H) in two densities. Afterwards, a representative volume element (RVE) for cement paste is developed in MATLAB based on the previous stage’s results (Stage 0). In the next stage, the RVE’s geometry is transferred to COMSOL Multi-physics via the Live Link™ tool for numerical analysis, where intrinsic mechanical properties are computed and verified with two independent experimental data sets. The verification showed that the proposed multi-stage model accurately predicts the compressive strength and Young’s modulus of cement paste with time. Therefore, this modelling approach can be extended to predict the material properties of mortar and concrete through upscaling.

PDF (Download Full text)

| Download Citation (XML)