3.1 Percolation and chemical shrinkage in cement pastes

  
Figure 1: Fraction connected porosity vs. total porosity for w/c=0.3 cement pastes of varying cement PSD and silica fume content. System size is 200 x 200 x 200 (0.5 µm/pixel).

This shift in porosity percolation threshold will also influence the empty porosity created by chemical shrinkage and the resultant autogeneous shrinkage occurring in water-cured cement pastes. Table 1 summarizes the empty and water-filled porosity present after executing 5000 cycles (about 25,000 hours or 1040 days) of the hydration model for both saturated/sealed and sealed curing conditions. The empty porosity due to chemical shrinkage for the finer cement is greater than that for the coarser cement for two reasons. First, the increased hydration of the finer cement results in an increase in chemical shrinkage, particularly evident in the systems hydrated under sealed conditions. Second, because the finer cement switches from saturated to sealed curing (when the capillary porosity depercolates) at a higher capillary porosity, it produces a greater amount of empty porosity during the "sealed" hydration stage. At a degree of hydration of 0.59, equivalent to that ultimately achieved by the 30 µm cement, the empty porosity volume fraction of the 5 µm cement is 0.024, 60 % higher than that observed for the coarser cement.

  

Figure 2: Original and final 100 x 100 2-D microstructure images for upper left: original w/c=0.3 PSD=5 µm, upper right: original w/c=0.3 PSD=30 µm, lower left: final w/c=0.3 PSD=5 µm, and lower right: final w/c=0.5 PSD=30 µm. Final images are for hydration under totally sealed conditions. In the original images, phases from brightest to darkest are: C₃S, C₂S, C₃A, C₄AF, hemihydrate, and porosity. In the final images, phases from brightest to darkest are: C-S-H gel, other cement, and empty porosity. Central bar extending across the microstructure is the flat plate aggregate.

In Table 1, the presence of 10 % silica fume is seen to result in a slight increase in the volume of empty porosity for both curing conditions, in agreement with previous simulation results [14]. While not directly addressed by the simulations, the cement PSD and inclusion of silica fume will also affect the pore size distribution of the hydrated cement paste. In two systems with equivalent total porosities and equivalent amounts of chemical shrinkage, the system with finer pores will exhibit a greater reduction in internal RH and a greater amount of autogeneous shrinkage [15]. In general, a reduction in empty porosity (due to chemical shrinkage) and resultant autogeneous deformation should be observed when using a more coarsely ground cement [10].