Next: Results Up: Main Previous: Introduction
All simulations presented in the results section were conducted using the NIST 3-D cement hydration and microstructural development model, which has been described previously [9]. The PSDs used were measured on actual cements [8] and bracket the PSDs currently produced by cement manufacturers. The fine cement had a median particle diameter of about 5 µm and a Blaine surface area of 640 m2/kg, while the coarse cement had a median particle diameter of about 30 µm and a Blaine surface area of 210 m2/kg. The composition of the cement determined by quantitative microscopy was 59 % C3S, 25.9 % C2S, 0.6 % C3A, and 14.2 % C4AF, with hemihydrate added at a mass percentage of 4.6 % [8]. Recently, a number of properties of these two model cements have been studied in detail [10]. Both cement paste and concrete (w/c=0.3) microstructures were simulated. In the latter case, the presence of an aggregate is simulated by placing a 2-pixel wide sheet of aggregate through the middle of the 3-D microstructure. For the 30 µm cement, simulations were also conducted with a 10 % silica fume replacement of cement (mass basis). The silica fume, modelled as 0.5 µm particles, reacts with the calcium hydroxide produced during hydration and also results in a reduction in the overall Ca/Si ratio of the C-S-H gel [11,12]. The cements were modelled as either 100 x 100 x 100 (1 µm/pixel) or 200 x 200 x 200 (0.5 µm/pixel) 3-D microstructures.
The hydration simulations were executed under two different curing conditions. In one case, the hydration was executed under totally sealed conditions, so that no additional water was available to replace that consumed due to chemical shrinkage. In this case, empty porosity is created within the microstructure to account for the water "volume" lost due to chemical shrinkage [9]. In the second case, referred to as saturated/sealed curing, the hydration was executed under saturated conditions until the capillary porosity became depercolated, at which point all subsequent hydration was performed under sealed conditions. Under saturated conditions, no empty porosity is created within the microstructure during hydration, as it is assumed that all needed water is readily available from the external environment to replace that "lost" via chemical shrinkage. Previously, Powers has suggested the moist curing of field concrete only to the point where the capillary porosity depercolates [13], as subsequent curing beyond this may be of little value.
Next: Results Up: Main Previous: Introduction