The results presented above clearly indicate that cement PSD has a large role in the autogenous properties of hydrating cement paste. A further consideration in most HPCs is the addition of mineral admixtures (pozzolans). For example, silica fume is composed of extremely fine (< 1 µm) particles that will finely subdivide the initial pore structure of the paste. This, along with the further increase in self desiccation caused by the pozzolanic reaction between silica fume and cement, will result in large reductions in RH and large increases in autogenous shrinkage [16]. Conversely, fly ash particles are typically of a size similar to that of cement and would be expected to have a smaller detrimental influence on autogenous properties. This is consistent with the results of Houk et al. [33], who studied a variety of concrete mixtures for the Dworshak Dam project and noted a "general increase in autogenous shrinkage with increase in fineness of the cementing materials.'' In their study [33], "concretes containing fly ash had cementing materials of lowest overall fineness and exhibited lowest autogenous shrinkage.".
The major problem in using coarser cements (and fly ashes) may be the lack
of early age strength development. In general, it appears that minimizing
cracking due to autogenous (and thermal) shrinkage/expansion and maximizing
early age strength are conflicting goals. If early age strength is not
an issue, the ideal concrete might be one with a relatively low cement
content [33], a low w/c ratio and a relatively coarse initial pore size
distribution that densifies (to nearly zero capillary porosity) due to
hydration. An alternate solution to the autogenous shrinkage and early age
cracking problem is the use of saturated lightweight aggregates to provide
autogenous curing for the lower w/c ratio concretes [34,35]. In
these systems, the water initially saturating the pores in the lightweight
aggregates is drawn into the hydrating cement paste by capillary forces to
provide the "extra" water needed to continue hydration and minimize
capillary stresses and autogenous deformation. Thus, the reduction in RH will
be controlled mainly by the size of the pores in the lightweight aggregates,
which will be the first to empty due to self desiccation. Here again,
strength will suffer somewhat due to the increased porosity present in the
aggregates. All of this demonstrates that the optimization of a high
performance concrete requires careful consideration of the end use performance
requirements, as well as the quality control achievable during field
production.