The results for the effect of the 90% RH environment on cement paste hydration are given in Fig 4. For the case of w/c=0.4, the 6 hour, 12 hour, and 3 day specimens show an effect due to hydration at 90% RH, while the 7 day specimen, over the first 28 days, does not appear to be affected by the change in environment. The data for the 0.3 w/c paste are similar, with the exception that the 3 day specimen also appears to be relatively unaffected by the change in environment.
These data, coupled with an understanding of the percolation of the capillary pore space, lead to a possible explanation for these effects. As the paste hydrates, the capillary porosity, and the characteristic capillary pore size, decrease. When the capillary porosity has been reduced to approximately 0.20, the capillary pore space is no longer interconnected throughout the paste [15,16]. When this occurs, vapor and moisture transport is restricted by the gel pore space, greatly reducing the transport coefficients. These gel pores, due to their small size (< 10 nm in diameter), will remain water-filled at RH values down to about 50% . Any capillary pores connected to the exterior environment only via a path containing gel pores will also remain water-filled at RH > 50%. Thus, when the capillary pore space is no longer percolated, the moisture is, in effect, trapped inside the specimen, and if the specimen is placed in an environment at less than 100% RH, but greater than 50% RH, the specimen continues to hydrate. The estimated degrees of hydration needed to achieve a capillary porosity of 20% are 0.63 and 0.43 for the 0.4 and the 0.3 w/c pastes, respectively, as illustrated by the horizontal dotted lines in Fig 4.
Figure 4a: Effect of a 90% RH environment on the degree of hydration () for an 0.4 w/c cement paste after curing in 100% RH for 6 and 12 hours, and 3 and 7 days. The solid line indicates continuous hydration at 100% RH and the horizontal dotted line indicates the degree of hydration at which the paste achieves 18% capillary porosity.
For the w/c=0.3 data, while the 7-day data appear to directly track the saturated condition hydration data, the 3-day data fall slightly below this curve. This might be expected, since after 3 days exposure to 100% RH, the 3-day specimen is then exposed to 90% RH, where even if the capillary porosity has depercolated so that no water is removed, the hydration is continuing under sealed, as opposed to saturated, conditions. As hydration slowly continues beyond 3 days, this will lead to the creation of some empty capillary pores in the specimen due to the chemical shrinkage. Thus, the measured degrees of hydration are seen to fall between the curves for hydration under totally saturated and totally sealed conditions. Conversely, for w/c=0.4, where the saturated and sealed curves will nearly overlap (according to the computer model results discussed earlier), the 3-day system is not able to achieve the performance of a sealed environment. In this case, 7 days of curing are required to equal the performance of a sealed system.
In terms of practical implications, it may be important to realize that when the capillary porosity depercolates, not only is it more difficult to remove water from the capillary pores in the specimen, but it is also more difficult to maintain saturation of the specimen via an external source of water. This effect has been observed during chemical shrinkage measurements on low w/c ratio cement pastes, where hydration as quantified by chemical shrinkage eventually falls below that measured by non-evaporable water content, as water can no longer be drawn into the specimen at a rate fast enough to maintain saturation [12,14,18]. This would imply that extended moist curing would be of limited benefit. However, moist curing during the first two to three days of hydration may be extremely important in maintaining internal saturated conditions to promote hydration, until depercolation of the capillary porosity is achieved. The situation will be even more complex in a concrete than in the cement paste specimens used in this study, due to the presence of the higher porosity interfacial transition zones surrounding each aggregate . It is possible that when the capillary porosity depercolates in the ``bulk'' paste regions of the concrete, it may still remain interconnected in the interfacial zone paste regions [20,21]. In this case, prolonged moist curing might be more valuable for a concrete than it would be for a cement paste specimen; more experimental measurements will be necessary to quantify the magnitude of this effect.