As many of the LTC scans have been presented elsewhere, 7, 24 here, only two representative series will be presented to introduce the subsequent analysis of percolated capillary porosity volume fraction. Figures 10 and 11 show a series of LTC scans after various hydration times for a w/c = 0.35 cement paste cured under saturated and sealed conditions, respectively. As mentioned in the Experimental section, the specimens cured under sealed conditions were resaturated (typically for one day) prior to the LTC scans. As many as three distinct peaks are observed in an individual LTC scan, corresponding to percolated capillary pores with a freezing (peak) temperature of about −15 ºC, percolated open gel pores freezing at about −25 ºC, and percolated dense gel pores freezing between −40 ºC and −45 ºC. 6 As mentioned earlier, the water frozen at −25 ºC and −40 ºC to −45 ºC may also include water in depercolated capillary pores that are connected together by open gel pores or dense gel pores, respectively. In Figure 10, as hydration progresses, the percolated capillary pores are observed to depercolate as the −15 ºC peak disappears between 3 d and 4 d hydration, while the open gel pores depercolate later, with the −25 ºC peak disappearing between 14 d and 30 d.
As shown in Figure 11, a substantially different behavior is observed for the specimens cured under sealed conditions and resaturated prior to the LTC scan. While the capillary pores in these systems do initially depercolate, they actually repercolate at ages of 14 d and beyond. Under sealed curing conditions, as chemical shrinkage and self-desiccation occur, shrinkage stresses and strains will be imposed on the C-S-H gel and its local shrinkage within a basically non-shrinking 3-D framework could reopen the originally closed capillary pore entryways.7 In addition, microcracking could also contribute to the detected percolated capillary porosity in Figure 11 at later ages. A similar repercolation of originally depercolated capillary pores has been observed by Bager and Sellevold for hydrated specimens exposed to drying followed by resaturation.14 This repercolation of a network of capillary (size) pores would be expected to have a large influence on transport properties (as indicated by the measurements of Powers and colleagues mentioned previously2) and durability in general. Particularly, the freeze-thaw performance of the specimens could be reduced as directly indicated by the higher freezable water contents for temperatures above −20 ºC for the 14 d to 50 d specimens. This observation reinforces the importance of avoiding this repercolation process by maintaining saturated curing conditions in low w/c cement-based materials, a condition that may sometimes only be assured by internal curing. 7
Similar LTC scans were obtained for the other w/c specimens and also for the 40 ºC curing.7, 24 One item of interest was that for the w/c = 0.35 specimens cured at 40 ºC, even though hydration is accelerated (as indicated by comparing Figures 3 and 4), it takes a longer period of curing time and a greater degree of hydration to achieve the initial depercolation of the capillary porosity, as indicated by the disappearance of the −15 ºC peak. In the 40 ºC cement paste, depercolation according to the LTC scans was observed to occur between 7 d and 14 d, as compared to the 3 d to 4 d observed above for the 20 ºC saturated curing. This observation is in agreement with the development of a coarser capillary pore structure when curing at higher temperatures, as mentioned above.17, 18
Figure 10. LTC scans for a w/c = 0.35 ordinary portland cement paste cured under saturated conditions at 20 ºC for the ages indicated in the legend.
Figure 11. LTC scans for a w/c = 0.35 ordinary portland cement paste cured under sealed conditions at 20 ºC for various ages indicated and then resaturated for one day.