6.1 Influence of Limestone on Hydration

Recently, the ASTM C 150 standard specification for portland cement has been modified to permit the use of up to 5 % by mass limestone in ordinary portland cements [40]. Much higher limestone contents have been routinely employed throughout the rest of the world. The CEMHYD3D v3.0 model can be conveniently applied to simulating the influence of such limestone additions on hydration (rates) and microstructure development [41]. When considering the influences of limestone on cement hydration, as introduced in section 4, two effects contribute to the observed influences: a fine particle effect where the limestone provides additional surfaces for the precipitation of hydration products, and a reactivity effect where the limestone participates in reactions with the calcium aluminate phases present in the hydrating cement paste microstructure. Generally, limestone is not very reactive, so that the influences are usually dominated by the first effect, that of a fine inert material.

Figures 4 and 5 present a comparison between CEMHYD3D v3.0 model predictions and measured experimental values for the degree of hydration of cement pastes with and without a 20 % by mass fraction substitution of limestone for portland cement. The simulations were conducted using starting cement paste microstructures that matched the particle size distributions (both cement and limestone), phase compositions, and phase distributions of the pastes investigated experimentally [41]. For all of the simulations conducted using CEMHYD3D v3.0, a conversion factor of 0.00035 h/cycle² was used to convert between model cycles and real time [41]. In the figures, the relative acceleration provided by the additional limestone surfaces is seen to be a strong function of the w/s of the cement paste. Thus, for w/s=0.435, little if any acceleration is observed when limestone is substituted for cement while at w/s=0.35, a considerable acceleration is observed. The substitution of limestone increases the effective w/c of the hydrating paste and this effect is much more important for w/s < 0.4, where the ultimate degree of hydration in the original cement paste is already limited to a value less than 1.0 due to space limitations (e.g., there is insufficient initial capillary pore space available for the precipitation of the volume of hydration products that would be created with complete hydration of the cement). Since the limestone is basically inert, its substitution at a constant w/s effectively provides a greater relative volume of pore space in which the hydration products can form. It must be recognized, however, that degree of hydration and strength development are not the same, as the projected increase in strength due to the acceleration "provided" by the limestone may be offset by the lower cement content (higher effective starting w/c) of the cement/limestone paste [41]. A quantitative analysis of these effects can be performed using the CEMHYD3D v3.0 model and its strength predictions based on Powers' gel-space ratio theory [39, 41].

Figure 4: Experimental and CEMHYD3D v3.0 model estimated degrees of hydration for CCRL cement 152 with and without 20 % by mass fraction limestone substitution for w/s=0.435, cured under saturated conditions at 20 ºC [41].

Figure 5: Experimental and CEMHYD3D v3.0 model estimated degrees of hydration for CCRL cement 152 with and without 20 % by mass fraction limestone substitution for w/s=0.35, cured under saturated conditions at 20 ºC [41].