The best validation of the 3-D microstructural model incorporating fly ash would be a comparison of the evolution of the model volumetric phase fractions over time with the equivalent experimental data. For crystalline phases, quantitative X-ray diffraction [11] would appear to be the most straightforward means for obtaining phase quantification. This would be particularly useful for the fly ash hydration products such as Freidel's salt and stratlingite. Alternately, SEM/image analysis may be useful in analyzing real microstructures at various hydration ages. Thermogravimetric analysis provides data on the water released at different temperatures, from which one can quantitatively estimate the CH content of cement/fly ash pastes. For example, this technique has been successfully used to determine the appropriate phase and volume stoichiometries for the reaction of silica fume with cement [12].
Ideally, these experimental studies should be conducted under isothermal conditions at several temperatures. This would enable the determination of activation energies and the variation of reaction products with temperature. These studies could then be extended to hydration under adiabatic conditions as presented in the previous section. If the phases forming at high temperatures are significantly different from those at lower temperatures, extensive modifications to the computer codes may be required. For example, ettringite is known to be unstable at temperatures above 70 º C [3], so that the model could be modified to make ettringite soluble whenever the system temperature exceeds 70 ºC, in addition to the current criteria for solubility based on the fraction of the initial gypsum remaining in the system.