Materials and Experimental Procedure

The Bogue composition of the cement employed in the experimental study, as determined by quantitative microscopy, is 59 % C₃ S, 25.9 % C₂S, 0.6 % C₃A, and 14.2 % C₄ AF. Rather than being interground with the cement, hemihydrate (C$\bar{S}$H_1/2) was added at a mass percentage of 5 %, just prior to mixing with water. This cement clinker was ground to four different finenesses corresponding to Blaine finenesses [11] of 643 m²/kg, 387 m² /kg, 254 m²/kg, and 212 m²/kg, respectively. The cumulative particle size distributions for these four cement powders, as determined by laser diffraction techniques are provided in Fig. 2. The corresponding average cement particle diameters obtained by fitting a Rosin-Rammler distribution [12] to the shown distributions are approximately 5 µm, 15 µm, 25 µm, and 30 µm, proceeding from left to right in Fig. 2. All mixes were prepared with freshly boiled demineralized water.

Figure 2: Particle size distribution for the cement powder ground to four different Blaine finenesses.

For this study, a w/c of 0.35 was chosen to be low enough to minimize bleeding, while also being high enough to avoid the necessity of using a water-reducing agent. For each mix, the cement powder was first dry-mixed with the hemihydrate addition. The appropriate mass of water was then added and the paste was mixed for 5 min in a five liter epicyclic standard mixer, with a brief pause after 2 min to scrape the sides of the bowl. The prepared paste was then cast in the holders for the various experimental measurements to be described in detail below. All samples were subsequently cured at 30 ºC. This curing temperature was chosen to enhance the thermal stability of the oil and water baths, and the environmental chamber used to hold the experimental equipment and specimens.