Performance Enhancement of Cementitious Materials by Nano-additives : An Experimental and Numerical Study

In recent years, nanomaterials like nano-silica and graphene oxide (GO) graphene oxide have been increasingly used to improve cementitious materials' mechanical and durability properties. However, the optimum dosage and achieving the desired mechanical properties of nano-modified cement-based requires significant time, effort, and cost. Furthermore, the precise reinforcing mechanism of these nanomaterials is not yet fully understood. Accurately predicting the dosage and mechanical properties of nano modified cement-based materials remains challenging. While earlier analytical models failed to account for the complexity of influencing factors accurately, newer Machine Learning (ML) techniques have shown promise but lack transparency. As the microstructure of hydrated cement paste significantly affects concrete strength, recent models have focused on simulating cementitious material hydration to bridge nano and macro properties. However, existing models often rely on assumptions and cannot directly correlate with concrete’s macro properties. Therefore, this research aims to explore the reinforcing mechanism of the nanomaterials and to develop a robust model capable of realistically predicting the mechanical properties of nano-modified cementitious materials by incorporating detailed microstructure information, fundamental physics and chemistry, thermodynamic considerations, and particle bonding and packing arrangements. Nanomaterials such as nano silica and GO were investigated for their impact on the mechanical properties of cementitious materials. Experimental studies, including Thermogravimetric Analysis (TGA), X-ray Diffraction (XRD), X-ray Fluorescence (XRF), Fourier-transform Infrared Spectroscopy (FTIR), Scanning Electron Microscopy (SEM), Silicon and Nuclear Magnetic Resonance (Si & H NMR), and zeta potential analysis, were conducted to understand the effects of these nanomaterials on formation of hydrates and evolution of microstructure. Nano silica enhanced the mechanical properties of cement paste, especially at early ages (< 7 days), through its pozzolanic and nucleation effects. However, an increase in nano silica content led to agglomeration, diminishing its effectiveness. Moreover, synthesizing calcium-silica-hydrate (C-S-H) with and without GO revealed that GO accelerated C-S-H sheet formation and promoted layer-by-layer C-S-H formation, resulting in very dense C-S-H capable of withstanding higher stress. This layer-by-layer formation also refined the pores. In addition, a robust chemical bond was observed between GO groups, such as carboxyl and Ca, which contain hydration products. A systematic model was then developed to optimize nano silica use in cement paste and to predict their nonlinear behaviour. The model integrates a hydration model, which calculates the dissolution rate of each clinker mineral, with a thermodynamic model simulating the hydration reaction and interaction between hydrates and nano-silica. The hydration model (HyMEC), which was developed in our laboratory, was modified to account for its reaction degree, dependent on portlandite availability, and to accurately represent the formation of C-S-H with a transition from jennite type to tobermorite C-S H. The model computed volume fractions of various phases, and a representative volume element (RVE) was formulated for cement pastes with and without nano silica using MATLAB. Finite element analysis with COMSOL Multiphysics was then used to evaluate the RVE, allowing computation of homogenized material properties such as compressive strength and Young's modulus. Predictions aligned well with experimental findings. Additionally, the model was extended to predict mortar mechanical properties, considering fine aggregate content and Interfacial Transition Zone (ITZ) properties, and validated with experimental results. In summary, this systematic model accurately predicts the nonlinear behaviour of cement pastes with different nano-silica replacement levels and can be extended to predict nano modified mortar, offering the potential for optimizing cement-based composites in various engineering applications.