 |
In low w/c (< 0.38) concretes, hydration of the cement will be incomplete, as there is insufficient pore space within which the hydration products can deposit. Thus, in these concretes, a portion of the relatively expensive cement is potentially being wasted and serving only as a reinforcing filler material. Recently, the VCCTL has been applied to investigating the potential of replacing a portion of the coarser cement particles by inert fillers [7]. A few example results from this study will be presented here.
As a base case, the study considered a concrete with a w/c of 0.25. All microstructures were based on the phase composition and PSD (provided in the cement images database) of Cement 135 issued by the Cement and Concrete Reference Laboratory in January of 2000, and previously characterized using the CEMHYD3D hydration model [6]. Initial microstructures were created with no cement replacement, and with the coarsest 20.5 % and 30.8 % of the cement particles (mass basis) replaced by an inert filler material. All three microstructures were created using Menu Selection 3 of the VCCTL. The inert particles were simply added by changing the phase ID for the appropriate larger diameter particles from cement to inert filler in the input table shown in Figure 7. Phases were then distributed amongst the cement particles in each system, according to the measured 2-D image values for Cement 135 [6], using Menu Selection 7. The resultant starting microstructures were then examined using Menu Selection 4, with representative 2-D slices being shown in Figures 23 and 24 for the systems with 0 % and 30.8 % cement replacement, respectively.
Figure 23: 2-D slice from 3-D starting microstructure for Cement 135, w/c=0.25, with no cement replacement by inert fillers.
Figure 24: 2-D slice from 3-D starting microstructure for Cement 135, w/c=0.25, with 30.8 % cement replacement by inert fillers.
The systems were then hydrated for 4000 dissolution/reaction cycles [2], representing about 200 d of real hydration time under saturated/sealed conditions at 25 ºC [6], using Menu Selection 13. Saturated/sealed conditions indicate that the hydration is initially executed under saturated (water is imbibed into the 3-D microstructure) conditions, but the curing is switched to sealed curing (no further water imbibition) when the water-filled porosity depercolates during the hydration process. For this study, the percolation characteristics of the water-filled capillary porosity were evaluated after every 100 cycles of the hydration process. Representative images of 2-D slices after the hydration are provided in Figures 25 and 26. In the latter image, the hydrated microstructure is seen to be uniformly dense, even in the immediate vicinity of the inert filler particles, due to the low overall w/s value.
Figure 25: 2-D slice from 3-D hydrated microstructure for Cement 135, w/c=0.25, with no cement replacement by inert fillers.
Figure 26: 2-D slice from 3-D hydrated microstructure for Cement 135, w/c=0.25, with 30.8 % cement replacement by inert fillers.
The predicted degrees of hydration and compressive strength developments are compared amongst the three systems in Figures 27 and 28, which provide comparisons of the base system (no replacement) vs. the systems with 20.5 % and 30.8 % replacement, respectively. Compressive strengths for ASTM C109 [14] mortar cubes have been estimated using Power's gel-space ratio theory [11], assuming that the inert filler particles and unreacted cement cores contribute equally to the compressive strengths of the mortar cubes. The systems with the inert filler replacements are observed to attain larger degrees of hydration, due to their effectively higher w/c ratio. While they have less cement reacting overall, this increase in hydration rate results in a predicted compressive strength development that is inferior to but on the same order as the system with no cement replacement. The actual decreases in predicted compressive strengths are plotted in Figure 29. For both replacement levels, the difference in compressive strength is observed to reach a maximum after about 14 d and then decay towards zero. While the > 10 MPa maximum predicted strength loss with the 30.8 % replacement level might be deemed unacceptable, the 6 MPa maximum loss with the 20.5 % replacement level is most likely a reasonable tradeoff for the cost savings that would result from replacing relatively expensive cement by a hopefully inexpensive basically inert filler (such as limestone).
Figure 27: Predicted degree of hydration and compressive strength development for base w/c=0.25 system and system with coarsest 20.5 % of cement particles replaced by inert filler. Solid line is original system and dotted line is system with 20.5 % replacement.
Figure 28: Predicted degree of hydration and compressive strength development for base w/c=0.25 system and system with coarsest 30.8 % of cement particles replaced by inert filler. Solid line is original system and dotted line is system with 30.8 % replacement.
Figure 29: Predicted reduction in compressive strength due to coarse cement particle replacement for Cement 135 with w/s=0.25. Solid line is reduction with 20.5 % replacement level and dotted line is reduction with replacement level of 30.8 %.
