The 4-parameter Compressible Packing Model (CPM) for sustainable concrete

International audienceTo reduce the environmental impact due to the CO 2-emission, it is necessary to optimize the concrete mix-design by reducing the amount of cement. In this case, analytical packing models are necessary to predict the packing density of a pile of grains. A new version of the Compressible Packing Model (CPM) (de Larrard et al.), the 4-parameter CPM, is introduced to predict the solid fraction of maximally dense disordered packings of bidisperse particles. It is apt to account for the loosening effect on big particles by interstitial small ones, and for the wall effect within assemblies of small particles near a big one, with moderate size ratios. The new theory is based on a specific treatment of configurations of one secondary class particle surrounded by dominant class neighbours. It has four parameters : wall effect and loosening effect coefficients, critical cavity size ratio and compaction index. The model proves its efficiency when compared to 780 results on various tested materials. The correlation coefficients between predicted and measured packing densities are very high: 99% for frictionless glass beads, 98,7% for spherical particles numerically simulated, 97,8% for natural aggregates and 96,4% for crushed aggregates. To predict the viscosity of the same grains in a high concentrated suspension compound of spherical, inert and rigid particles suspended in a Newtonian fluid, we resort to the iterative approach advocated by Farris and to the Krieger-Dougherty power-law relation for the relative viscosity. The theory was developed to highlight a new relation between relative viscosity and the solid volume fraction, compatible with Einstein's relation