Several previous studies have shown a quantitative relationship between setting as measured by the Vicat needle method [14] and the percolation of the solids in a three-dimensional microstructural model [19, 20]. For the standard ASTM technique [14], the Vicat measurements are generally made on a rather low (< 0.3) w/c cement paste. It is expected that the setting process would be a strong function of w/c [15-17], as the bridges being constructed to connect the particles together will generally need to be longer and/or fewer bridges per unit hydration of cement will be created when the w/c is increased.
Experimental and computer modeling results examining four different views of the "setting" process are provided in Figures 2 and 3. Figure 2 compares experimental measurements of the yield stress via stress growth measurements [15-17] with the computer modeled volume fractions of total solids, both as a function of hydration time for cement pastes with three different w/c. As the cement particles flocculate together following mixing, they form a weak solid skeleton that is strengthened and reinforced by the cement hydration products. For a high enough w/c, this initial skeleton can not support itself under gravity so that settlement and bleeding will occur. In fact, minor bleeding was observed for the w/c=0.45 cement paste prepared with CCRL cement 152, and the initial yield stress values in Figure 2 are quite low. As the w/c is lowered to 0.4 and then to 0.35, the measured yield stress becomes higher as a greater applied stress is necessary to get the (greater number of) particles moving in the more concentrated suspensions. Initially, the projected solids volume fraction and the measured yield tress track each other fairly closely, but for each w/c ratio, a point is reached where the measured yield stress climbs rapidly while the modeled volume fraction of solids continues to increase at a relatively constant and much slower rate. It is here that the percolation of the partially hydrated cement particles by the hydration products comes into play.
Figure 2: Two "views" of the setting process in cement paste as a function of w/c: yield stress by stress growth measurements (top) and total solids volume fraction (bottom).
The bridges built by the hydration products are much stronger than the initial interparticle forces flocculating the particles together. As these bridges percolate the three-dimensional microstructure, a finite resistance to the penetration of the Vicat needle is developed, as shown in Figure 3. The time at which the needle resistance, equal to (40 mm − the needle penetration in mm), begins to increase from zero in Figure 3 is seen to correspond closely to the time when the measured yield stress begins to diverge in Figure 2. In Figure 3, the needle resistance measurements are seen to also be in agreement with the modeled volume fraction of connected solids that characterizes the volume fraction of unhydrated cement clinker particles that are bridged by hydration products. All of this has occurred during the time when only the first 4 % to 8 % of the cement has hydrated. Already at this point, the solid skeleton is well in place and the strength of the microstructure will continuously increase as new bridges are formed and existing ones are expanded.
Figure 3: Two more "views" of the setting process in cement paste as a function of w/c: needle resistance (top) and connected solids volume fraction (bottom).