The measured chemical shrinkages for the two cement pastes are provided in Fig. 2. From the figure, it can be seen that the limestone is basically functioning as a simple dilutant in terms of the observed chemical shrinkage, as the normalized chemical shrinkage for the blended system is basically identical to that of the original cement. No indication of a significant acceleration or retardation or significant reactivity of the limestone is indicated by the chemical shrinkage results. This is consistent with the majority of the results present in the literature as summarized by Hawkins et al. . As shown in Fig. 3, the CEMHYD3D model provides a good fit to the experimental chemical shrinkage data for the blended system. A similarly good fit (not shown) was obtained for the base Cement 135 system.
The compressive strength results are provided in Fig. 4. The CEMHYD3D predicted values for the 7 d strength lie well within the precision of the experimentally measured values. At 56 d, there is more deviation between model and experiment, but more importantly, experimentally, there is no observed strength difference between the original and the blended systems, as both achieved average compressive strengths of nearly 100 MPa. The extra hydration achieved in the blended system at later ages, due to its effectively higher w/c ratio, appears to be sufficient to compensate for the reduction in cement content [2,4]. It would be expected that the two systems have basically equivalent amounts of capillary porosity at later ages (and thus equivalent strengths according to Power's gel-space ratio theory).
It is important to note the criticality of replacing the coarse cement particles with limestone. If one were to replace the finer cement particles instead, the computer simulation results suggest that a significant reduction in achieved hydration and compressive strength development would be obtained in the equivalent (limestone volume fraction of 15 %) blended system. Naturally, this is because it is the finer cement particles that make the largest contribution to hydration and strength development, particularly at early ages [1,10]. Intergrinding of the limestone and cement will likely arrive at an intermediate between the coarse limestone/fine cement and fine limestone/coarse cement systems. The studies conducted by Bonavetti et al.  have in fact indicated that the higher fineness often produced in interground limestone blended cements, along with their higher effective w/c ratio, may result in systems whose performance in terms of hydration and strength is basically equivalent to the original (lower fineness) cement. Whether the limestone is interground or directly replaces the coarser cement particles, the usage of cement is reduced, with concurrent ecological and economical advantages. However, the advantage of classification and replacement over intergrinding could be in a cost and energy savings, since no energy will be expended in further grinding of the coarse limestone.