Specimen Preparation

The cement pastes analyzed in this study were prepared using Cement and Concrete Reference Laboratory (CCRL) portland cement proficiency sample 152, ¹⁶ issued in January 2004. Its particle size distribution, as measured by laser diffraction techniques on a dispersion of the cement powder in isopropyl alcohol, is provided in Fig. 1 and its phase composition as determined by SEM image analysis is given in Table 1. In addition, the cement contained 6 % calcium sulfates by volume, distributed as approximately 44 % gypsum (calcium sulfate dihydrate), 52 % hemihydrate, and 4 % anhydrite, as determined by x-ray diffraction measurements. Cement pastes with initial w/c=0.35 and w/c=0.45 on a mass basis were mixed for several minutes at 20 ºC in a temperature-controlled high speed blender. Cast wafers (≈5 g) of the prepared pastes, approximately 32 mm in diameter and 2 mm to 5 mm in thickness, were placed in small, capped plastic vials to be cured at 20 ºC. After about 4 h of curing, any accumulated bleed water was removed from the vials using a pipette, to assess the "true" effective w/c of the pastes. The w/c=0.35 pastes contained negligible bleed water, but for the w/c=0.45 pastes, after removing the accumulated bleed water, a paste with an effective w/c=0.435 remained.

Three curing conditions were employed. In saturated curing, a small amount of distilled water was placed on top of the paste wafers after removing the bleed water. In sealed curing, the wafers were simply sealed in their plastic vials after removal of the bleed water. In sealed/saturated curing, the wafers were cured under sealed conditions for 7 d, then the plastic vials were opened and a small amount of distilled water was added on top of the wafers. All curing was conducted inside of a walk-in environmental chamber maintained at 20 ºC. At various ages between 1 d and 256 d, wafers were removed from their vials and crushed into small pieces. Often, some of the small pieces were retained for SEM or LTC analyses as described below. The remaining pieces were further crushed to a fine powder, flushed with methanol in a thistle tube under vacuum to stop their hydration, and divided between two crucibles (for replicates) for determination of degree of hydration via LOI measurements.

Table 1. Measured area (volume) and perimeter (surface area) fractions for CCRL cement 152. Numbers in parentheses indicate one standard deviation among a set of six SEM X-ray map images, each image having a 256 µm by 200 µm field of view. ¹⁷