Water layer thickness of silica fines and their effect on the workability of cement pastes

Concrete is used in infrastructure and in buildings. It is composed of granular materials of different sizes and the grading of the composed solid mix covers a wide range. The overall grading of the mix, containing particles from 300 nm to 32 mm, determines the mix properties of the concrete. The properties in fresh state (flow properties and workability) are for instance governed by the particle size distribution (PSD) and the resulting particle packing (PP). One way to further improve the packing is to increase the solid size range, e.g. by including particles with sizes below 300 nm. Possible materials, which are currently available, are limestone and silica fines like silica fume (mS) and nano-silica (nS). This paper addresses the characterization of six different silica fines with respect to their application in cement paste. Given that the fines provide by far the highest percentage of specific surface area in a mix, their packing behavior and water demand is of vital interest for the design of concrete. In the present work, different mixes are compared and analyzed using the mini spread-flow test method. In this way, a deformation coefficient derived by the spread-flow test is confirmed to correlate with the product of computed specific surface area (SSA) based on measured PSD and intrinsic density of the individual silica fines. Similarly, correlations with equal accuracy are found with a computed SSA using the BET method. With the flow experiments of different mixes it is possible to derive an individual deformation coefficient of the silica particles. It is demonstrated that the computed and the BET surface area values have a constant ratio (0.76 to 0.70). Finally, the value of a constant water layer thickness around the powder particles (24.8 nm) is computed for all silica fines at the onset of flowing. This implies the possibility to predict the flow behavior of paste only based on the knowledge of their SSA, either determined by computation or by BET measurements.