Figures 8.4 and 8.5 present for each of the paste mixtures included in Table 8.3 the resulting values of total voids u for varying VJVcm ratios. Optimum packing points for each paste mixture are also noted on the u versus Vw/Vcm curves (greyfilled squares). The water voids line is drawn in each of the Figs. 8.4 and 8.5 graphs, in order to facilitate the estimation of airfilled voids content of every single paste mixture. Watertocementitious materials ratios and solid concentration factors at MPP and OPP for all paste mixtures are listed in Table 8.4.
From Fig. 8.4a and Table 8.4 it becomes clear that the increase of SP dosage in cementonly pastes (without additions) allows for the reduction of Vw/Vcm, leading to the decrease of the total voids content (or, equivalently, to the increase of the mixture’s packing density), while maintaining the workability at a specified level
Table 8.4 Fresh — and hardenedstate results of all produced cement paste mixtures

(that of the OPP — see greyfilled squares in Fig. 8.4a). Even when the same Vw/Vcm ratio is used, higher SP content results in lower voids ratios u and, hence, in higher solid concentration indices, f; this observation comes into agreement with the findings of (§hamaran et al. 2006; Jones et al. 2003).
The effect of additives on the packing density of SA pastes is more pronounced, especially for finelygrained ones. The addition of LF120 (Fig. 8.4b), as cement replacement (by volume), produces pastes with OPPs that correspond to lower Vw/Vcm ratios and higher packing densities compared to additionfree pastes (CEM15%sp). With varying percentage of cement replacement by LF120 (up to 25% by volume) there appears to be a linear relationship between Vw/Vcm and u (hence, f) at the optimum packing points. However, for cement replacement percentages larger than 25%, voids volume decreases marginally (for any given
V /V ratio) and packing at OPP reaches a maximum attainable value. For the same Vw/Vcm ratio, the increase of the packing density of a paste with increasing quantity of cement replaced by a coarser powder comprises — by intuition — a contradiction. Nevertheless, the gradual substitution of cement with LF120 results in the modification of the solid particles’ grading curve so that better packing is achieved.
In regards to LF120, the effect of cement replacement by the finer LF10 limestone powder was qualitatively the same but more noticeable in terms of packing density increase at the OPPs with increasing cement replacement ratios (see Fig. 8.4c). In this case, a maximum attainable value of packing density at OPP was not reached, even for cement replacement percentages as high as 50%.
Cement replacement by silica fume resulted in considerable increase of the pastes’ packing densities at MPPs (Fig. 8.4d), whereas the respective increase at OPPs was of lower magnitude. This was owed to the fact that increase of cement replacement percentages by silica fume resulted in pastes of progressively higher cohesion and stickiness. This was especially noticed for a cement replacement percentage equal to 50% (a percentage not practically applicable from a commercial point of view); in this case, the Vw/Vcm ratio corresponding to the OPP was equal to 0.72 (i. e. higher than the ones at OPPs for SF125% and SF25% pastes and considerably higher than the one corresponding to the MPP of SF50%).
The first set of MA pastes comprising limestone fillers, showed limited capacity for packing density increase at the OPPs for increasing content of additive LF10 up to 11.5% by cement volume. For these mixtures optimum packing density was achieved at an approximately constant Vw/Vcm ratio equal to 0.66 (see Fig. 8.5a and Table 8.4). The paste with LF10 content equal to 17% by cement volume (15% by cement weight) exhibited the highest packing density. The effect of SF content increase in MA pastes comprising LF120 and SF was analogous to the one for (SA) SFY% pastes. Again, with increasing SF content, packing density at OPPs increased at a much lesser extend in regards to the achieved increase of packing density at MPPs. As in the case of LF120 + LF10 Y% pastes (where 6% < Y% < 11.5%), optimum packing density in the LF120 + SF Y% ones (where 7% < Y% < 14%) was achieved at the same V /V ratio (with a mean value of 0.66). Further increase in SF content (LF120 + SF215%) resulted in a dramatic decrease of packing density at the OPP. A final MA paste mixture comprising finely ground powders (LF10 + SF11%) was investigated (Table 8.4); this mixture resulted in the maximum achieved packing density at the OPP compared to all other tested paste mixtures (and at a value of
V /V ratio lower than 0.66).
w cm