In addition to the new features and enhancements outlined in the previous five subsections, several miscellaneous additions have been made to CEMHYD3D v 3.0 computer codes. First, the ability to begin hydration under sealed conditions and later switch to saturated conditions has been included by allowing the user to specify a number of cycles of hydration after which the capillary porosity is once again filled (saturated) with water. Experimental studies have indicated that for intermediate w/c (0.4 to 0.45), some combination of sealed followed by saturated curing may result in an earlier depercolation of the capillary porosity, as hydration will initially be somewhat concentrated in the pore entryways and smaller pores, and not in the larger pores that will be emptying due to self-desiccation during sealed curing [25].
Second, the user has the option to specify the preferred morphology of the C-S-H gel formed during the hydration as either "random" or "plates". In the presence of sufficient alkalis, SEM evidence indicates that plate (lath-like) formations of the C-S-H gel may form as opposed to the more random morphologies observed in low-alkali cement pastes [26, 27]. The formation of a plate vs. a random morphology hydration product may result in an earlier depercolation of the capillary porosity, as exemplified by recent low temperature calorimetry measurements on a low-alkali cement with and without additional sodium and potassium sources (either hydroxides or sulfates) [28]. With this option, the assumption is being made that the morphological changes observed by SEM at scales of tens and hundreds of nanometers are also present at the one micrometer and higher scale implemented in the CEMHYD3D v3.0 model; this would infer a degree of self-similarity to the overall cement paste microstructure [29].
Finally, the user has the option of specifying a special dissolution bias factor for the one-voxel particles in the hydrating microstructure. Bullard has recently presented a derivation relating this factor to the PSD of particles smaller than a certain size (e.g., 2 μm) [30]. The basic premise is that in cements with a significant fraction of thei particles smaller than 1 μm in diameter, representing the particles as 1-voxel (1 μm) particles in the CEMHYD3D model may result in the initial hydration rates being much lower than those observed experimentally (due to surface area effects, for instance). Via this factor, the user may increase (or decrease) the dissolution rates of (only) the one-voxel particles to achieve better agreement between model and experiment.