A plot of shear wave energy obtained from ultrasonic experiments on a set of cement slurries is provided in Figure 18, from D'Angelo et al (1992). This figure shows a sharp onset, where significant shear wave energy is first transmitted, which we interpret as the point at which a continuous solid backbone first exists, as liquids cannot sustain a shear sound wave in the low-frequency limit. After this initial onset, the amount of energy transmitted rises sharply as the connected backbone quickly assimilates the remaining isolated solid clusters. There is some experimental noise in the steeply ascending part of this plot, which may or may not have a microstructural origin. When all solids are part of the connected backbone, the shear modulus will then continue rising more gradually as the porosity continues to decrease with hydration. The similarity of the solids percolation plot in Figure 15 to the shear wave plot in Figure 18 is striking. This agreement is further evidence that the percolation arguments suggested by the model are the correct explanation of the shear wave data.
The degree of hydration at solid percolation, as predicted by the model, varies with W/C ratio and particle size distribution. However, for W/C ratios near those employed in the experiments shown in Figure 18, the range of variation in degree of hydration is not great, from about 1 to 10%. There are no other adjustable parameters in the model, since the amount of products formed per unit volume of cement hydrated is fixed by the chemical reactions used, while the various diffusion parameters associated with CH formation do not affect the solid percolation (cement + C-S-H) (Chiesi et al 1992, Bentz and Garboczi 1991b). The onset of shear wave energy transmission in Figure 18 clearly occurs between 2-6h of hydration time. These times are perfectly consistent with degrees of hydration in the 1-10% range, lending support to our explanation of Figure 18 in terms of a percolation transition.