Fractal-based microstructure reconstruction to predict the permeability of cement pastes

The permeability of cementitious materials is one of the crucial indicators in the quantitative evaluation of durability, and it is significantly affected by the characteristics of the internal pore structure of such materials. To predict the permeability of cementitious materials based on a clear analysis of their pore structure, we propose an improved methodology for virtually reconstructing the three-dimensional pore structures of ordinary Portland cement (OPC) pastes by using fractal theory. In addition, we propose a methodology to predict permeability on the basis of the reconstructed pore structure. The accuracy of pore structure reconstruction was found to be the highest when each pore region was reconstructed using five types of base fractal units (BFUs) based on the fractal dimensions of pore-size distribution and pore tortuousness. The permeability predictions generated using the proposed methodology indicated that when the volumetric parameter of the pore structure was the same, the permeability increased by up to four orders of magnitude as the geometric parameters varied. In addition, the fractal dimension of pore-size distribution had a more critical effect on the large capillary pores region, and the fractal dimension of pore tortuousness had a more critical effect on the gel & small capillary pores region. These results indicate that consideration of the geometric characteristics of the pore structure significantly influences the pore structure reconstruction and permeability prediction processes.