 |
The heats of hydration of the two cements were assessed using a multi-chambered microcalorimeter constructed at NIST [10]. A known mass of cement, along with several small stainless steel balls to facilitate mixing, were placed in a sealed calorimetric cell which was then equilibrated in the main calorimeter chamber. After a steady heat flux signal was obtained, the cell was removed, the appropriate mass of water (also thermally equilibrated to the calorimeter temperature) quickly added using a syringe, and hand mixing performed (by shaking the cell) before restoring the cell to the calorimeter chamber. The voltage signals produced (proportional to heat flux) by the calorimeter cells were digitized using a PC-based high resolution A/D data acquisition system. Thus, during the initial hydration, data could be taken at 30 second intervals. Once the reactions slowed, data was typically acquired every 5 or 10 minutes over a period of at least 7 days. At longer times, the signal of the calorimeter is very close to its background level, so that detection of the slow but ongoing hydration becomes unreliable. In analyzing the heat release data, the initial exothermic "mixing" peak was ignored due to the necessity of removing the sample cell from the calorimetric chamber to assure adequate mixing. This could result in a difference in the cumulative heat released over a period of 7 days on the order of 10 kJ/kg or about 4%, as estimated from samples mixed insitu in the calorimeter. Due to the mixing difficulties at low w/c ratios, calorimetric measurements were only performed at the two higher w/c ratios of 0.4 and 0.45.
Figure 4 provides a sample plot of the obtained signal for Cement 116 at 25 ºC for w/c=0.4 for the first 24 hours of data acquisition. This signal vs. time was then numerically integrated to obtain the cumulative heat release (kJ/kg cement) vs. time curves which will be presented in the results.