Studies were conducted utilizing Cement and Concrete Reference Laboratory cement sample 135 [5] and a commercially available white limestone powder obtained from OMYA, Inc. 2 The two powders were particle size separated using an Alpine air classifier with a cutoff size of approximately 30 µm. The fine fraction of the cement (predominantly below 30 µm) was blended with the coarse fraction of the limestone (predominantly above 30 µm) in a v-blender to prepare a blended system with 15 % limestone by volume (assuming specific gravities of 2.71 for the limestone and 3.2 for the cement). A 15 % by volume replacement level was chosen based on previous computer simulations [2] that suggested this to be the maximum replacement level for a w/c=0.3 system without compromising performance. The particle size distributions (PSDs) of the original and fine cements, the coarse limestone, and the blended system were measured using laser diffraction techniques. The cumulative PSDs so obtained are shown in Fig. 1, along with the calculated distribution for the blended system based on its volumetric/mass proportions. It can be observed in Fig. 1 that the separation was basically successful, with only 10 % of the particles being finer or coarser than the 30 µm cutoff for the limestone and cement powders, respectively.
The blended system was used to prepare cement pastes and mortars with a water-to-solids ratio (w/s) of 0.3. Control mixtures were also prepared using the original cement 135 powder. For the pastes, to assess hydration rates, chemical shrinkage measurements [6] were executed for a period of 10 d. The maximum expanded uncertainty [7] in the calculated chemical shrinkage has been previously estimated [6] to be 0.001 ml/g, assuming a coverage factor of 2 [7]. For the mortars, compressive strength cubes were prepared according to ASTM C109 specifications [3], but with a w/s=0.3. Their compressive strengths were evaluated after 7 d and 56 d of curing in a tank of lime-saturated water. Three cubes were tested for each mixture at each age. The mixture proportions for the mortars are provided in Table 1. All curing and measurements were conducted at a temperature of 25 ºC.
| Table 1:Mixture proportions for mortars for strength testing | |
|---|---|
| Material | Mass (g) |
| Cement 135 or blended system | 1035.6 |
| Water | 301.2 |
| Graded sand | 1910.6 |
| Water-reducing admixture (ASTM C494 Type A [3] (naphthalene sulfonate-based)) |
15.69 |
All simulations were conducted using version 2.0 of the NIST CEMHYD3D program [6,8] and the Virtual Cement and Concrete Testing Laboratory web-based interface [9]. Three-dimensional starting microstructures with w/s=0.3 were created based on the measured PSDs and phase compositions of Cement 135 [5] and the blended system. The starting microstructures were then hydrated using the CEMHYD3D codes and the calculated chemical shrinkage and compressive strength developments compared to the experimental data. In the model, compressive strengths were calculated based on Power's gel-space ratio concept [6,10] with a strength prefactor of 123.5 MPa for the mortar cubes. Power's gel-space ratio for strength prediction has been applied successfully in the past to limestone blended cements [4]. While limestone is not extremely reactive, its slow conversion to a monocarboaluminate phase (AFmc- (CaO)3(Al2O3)-CaCO3-11H2O) [11,12] was included in the reactions present in the updated CEMHYD3D codes.