Hydration Kinetics

Figures 9 and 10 show the model predictions and experimental results for degree of hydration under saturated and sealed conditions for cements 135 and 136, respectively. In both cases, CEMHYD3D does an excellent job of fitting the experimental data and in matching the observed differences between saturated and sealed curing observed at longer hydration times. Early in the hydration process, excess water is readily available, so that the water lost due to self-desiccation in the sealed specimens does not have a major influence on hydration kinetics. As the specimen ages, the amount of self-desiccated (empty) porosity becomes significant, and the hydration rate of the sealed specimens trails behind those of the corresponding saturated specimens. While not a large difference is observed for the w/c=0.4 samples investigated in this study, for lower w/c ratios (now often being employed in high-performance concretes) sealed conditions can lead to a substantial reduction in the degree of hydration (and the strength) achieved at long hydration times (e.g., > 7 d). For these two cements, the conversion factor between hydration cycles and real time is nearly a constant (0.00030 for cement 135 and 0.00036 for cement 136). These values are in good agreement with a previously used conversion factor of 0.00030 for cement 133 issued by the CCRL in 1999 [4].

  

Figure 9: Computer model (lines) and experimental results (data points) for degree of hydration of CCRL cement 135, w/c=0.40, under saturated and sealed curing conditions. Crosses indicate ± one standard deviation in experimental measurements, and generally fall within the boundaries of the data point symbol itself. Model cycle to actual hydration time conversion factor was 0.0003.

  

Figure 10: Computer model (lines) and experimental results (data points) for degree of hydration of CCRL cement 136, w/c=0.40, under saturated and sealed curing conditions. Crosses indicate ± one standard deviation in experimental measurements, and generally fall within the boundaries of the data point itself. Model cycle to actual hydration time conversion factor was 0.00036.