In the past, for a given cement, the cycle-to-time conversion factor in CEMHYD3D has been adjusted to provide the best fit to measured degree of hydration data for the cement as a whole, usually based on loss-on-ignition (LOI) measurements. The data set of Gutteridge and Dalziel [4] offers the possibility to examine and adjust the individual phase reactivities. They prepared pastes with a water-solids mass ratio (w/s) of 0.71 and monitored the consumption of the four major cement clinker phases using quantitative X-ray diffraction from ages of 6 h to 180 d. Their cement had a mass composition of 68 % C3S, 14 % C2S, 6 % C3A, and 7 % C4AF, with 2 % calcium sulfate. Its density was 3160 kg/m3 and its fineness was 320 m2/kg. For modeling in CEMHYD3D, it was assumed that the individual phase surface area fractions were equal to their bulk volume fractions. Then, a cement in the cement images database with a similar composition and fineness was selected to provide the needed correlation files. Model hydration was conducted under sealed conditions to mimic the experimental procedure and the individual phase volume fractions were monitored as a function of hydration cycles. A conversion factor of 0.00045 h/cycle2 was used to convert between hydration cycles and real time, to provide a good fit to the overall degree of hydration data (based on LOI) provided by Gutteridge and Dalziel [4]. In comparing the degree of hydration data for the individual phases between the experiment and model, the fits for tricalcium silicate and tricalcium aluminate were quite reasonable without any further adjustment from version 2.0 of CEMHYD3D. But, it was observed that the dicalcium silicate and tetracalcium aluminoferrite were too reactive in version 2.0 of CEMHYD3D. Thus, their individual phase dissolution probabilities were each reduced by 50 % from their values in version 2.0. As can be seen in Figure 2, this adjustment resulted in reasonable fits for all four sets of the degree of hydration data for the individual cement clinker phases.
Figure 2: Comparison of CEMHYD3d v3.0 degrees of hydration for individual cement clinker pahses with experimental data of Gutteridge and Dalziel [4].