Effect of hydration on the value of D_ITZ/D _bulk

An interesting prediction of the multi-scale diffusivity model is that the contrast in diffusivity, D_ITZ /D_bulk, between the ITZ and the bulk cement paste, changes with hydration. This is easily seen for a homogeneous C₃S cement paste, for which the capillary porosity $\phi$ as a function of degree of hydration $\alpha$ is given by the straight line [8]:

$$\phi = 1 - \frac{1+1.13 \alpha}{1 + 3.21 w/c}$$ (24)

The value 3.21 is the specific gravity of C₃ S, and the value 1.13 comes from the volumes of various hydration products allowing for chemical shrinkage and fully saturated curing conditions [8]. For different values of w/c, the slope of the line is different. For the system considered, the porosity in the ITZ and the bulk regions was found to decrease linearly with degree of hydration. Since the cement paste diffusivity depends non-linearly on porosity, that means that the diffusivity ratio between bulk and ITZ paste will change with hydration. Figure 17 shows the results for D_ITZ/D _bulk for this system as a function of degree of hydration. The points are from recent experimental results [60], obtained by applying the D-EMT to the experimental measurements. There is qualitative agreement between experiment and the multi-scale model predictions. The degree of hydration at which the maximum value of D_ITZ/D _bulk is obtained is about 0.6 for experiment, and 0.7 for the multi-scale model. Also, the experimental points, except at late hydration, are well above the model points. This implies that the model needs improvement for handling how much hydration product forms in the ITZ, in order to make the porosity difference greater. One should also note that the experimental points were obtained by applying D-EMT to experimental data, and extracting the effective value of D_ITZ/D _bulk, so that there is probably some systematic error in the experimental points. Notice that at later hydration, the agreement becomes closer, implying that the multi-scale model gives more quantitative predictions for later hydration times. For durability considerations, it may only be necessary to give accurate diffusivity predictions at later hydration times.

Figure 17: Plot of the value of D_ITZ/D_bulk vs. degree of hydration for the example mortar discussed in the text.