An advanced multiscale analytical approach for predicting the Equation of State (EOS) of fully saturated cement pastes is presented. This approach models the matrix as an elastic–plastic material with linear hardening that includes capillary pores filled with water. The model takes into account both the stochastic variation of pore sizes and their spatial distribution. The microscale level is represented by a spherical medium having the mechanical properties of the cement paste matrix. A single spherical pore that is filled with compressible water is located at the center of the micro-domain. The behavior of the compressible water is modeled by the nonlinear Tait equation of state. The EOS at the macro level is obtained by averaging the micro level domain strains over the pore sizes and pore wall thicknesses. It is shown that the solution depends only on the ratio between the minimum pore wall thickness and the lower limit of capillary pore radius and is independent of each of these parameters separately. The EOS obtained by the proposed model shows good agreement with experimental data for cement paste with water/cement ratio w/c = 0.50. The performed parametric study shows that the bulk modulus and yield stress of undisturbed cement gel matrix significantly affect the EOS, while the variation of Poison ratio, plastic hardening parameter and minimum limit pore size and wall thickness (within physically accepted range) slightly affect the cement paste bulk behavior.
Read full abstract