In addition to the data available from the CCRL proficiency sample testing program [4,5], supplementary experimental measurements were performed at NIST. These included the determination of the non-evaporable water content of hydrating cement pastes to assess degree of hydration as a function of age (as described in detail in reference [12]), chemical shrinkage [9], and pore solution composition. The latter analysis was performed using ion chromatographic analysis of pore solutions extracted from the hydrating cement pastes after various ages. For early ages on the order of a few hours, the pore solution was obtained simply by filtering about 150 g of hydrating cement paste (w/c=0.4). For later ages, it was necessary to express the pore solution from the hardened cement paste using a special die and a universal testing machine, based on the methodology developed by Longuet [13]. For three replicates analyzed by ion chromatography, the maximum standard deviation in the determined concentrations was less than 0.004 mol/L.
All simulations were performed using the VCCTL software. Starting images of CCRL cements 135 and 141 at the necessary water-to-cement (w/c) ratios (0.25 to 0.485) were created and hydrated for various amounts of time, depending on the specific property of interest (e.g., > 28 d for compressive strength, about 10 h for setting time comparisons). As hydration occurs in the VCCTL system, a set of output files are continuously updated to reflect the development of properties with time. These results can be easily compared to those available either from the CCRL proficiency sample testing program or the experimental measurements performed at NIST. Details on the cement hydration and microstructure development computer model used in the VCCTL can be found in references [9] and [10].