Electrical properties of cement-based materials: Multiscale modeling and quantification of the variability

Abstract The electrical properties, resistivity and conductivity, inform on the durability of cement-based materials and can be used for monitoring and inspection of concrete structures. The physical origin of these properties can be linked to the dynamics of the pore solution. We propose a multiscale modeling approach of the electrical conductivity and resistivity informed by the dynamics of ions that enables the quantification of property variability across scales using Monte Carlo Micromechanics (MCM) computations. As a source of variability, we consider the pore solution composition, clinker composition and the uncertainty on solid conductivity. The results are compared to experimental measurements on various cement systems. The age-dependency of ionic diffusion, due to ion-ion and ion-solvent collective effects, is crucial to model the evolution of electrical conductivity. The main results show that Monte Carlo Micromechanics enables the quantification of the variability and uncertainty across scales, since MCM computations have provided estimates of the standard deviation of the electrical conductivity and resistivity at the scales of cement paste, mortar and concrete. Also, the results show that self-consistent scheme provides a good estimate of the effective electrical conductivity of cement-pastes, capturing the transition from a liquid to a solid matrix during cement hydration.