Next: Computer Modelling Techniques Up: Main Previous: Main
In 1935, Powers [1] reported that during cement hydration, external water is absorbed by the hydrating cement paste to replace that consumed by chemical shrinkage (the hydration products occupying a smaller volume than the reactants). When external water is not present or not readily available to the hydrating system, self-desiccation occurs and empty pores are created within the microstructure. Based on the Kelvin-Laplace equation, these empty pores will cause a reduction in the internal relative humidity (RH) [2], in turn inducing capillary stresses within the remaining capillary water. These stresses will cause a measurable deformation, referred to as the autogeneous shrinkage of the mortar or concrete specimen.
The magnitude of autogeneous shrinkage in conventional concrete mixtures is very small, but increases dramatically as the water-to-cement (w/c) ratio decreases to the values used in high-performance concretes.
The reduction in internal RH is mainly dependent on two factors: the volume of empty pore space created and the pore size distribution of the cement paste [3]. Similar to a mercury intrusion experiment where the largest connected pores first fill with mercury, during self-desiccation, the largest pores will empty first. Thus, the moisture content vs. relative humidity tends to follow the measured desorption isotherm of the material [4]. When attempting to engineer the autogeneous shrinkage and internal RH of a concrete for field use, one can attempt to regulate either the amount of empty pore space, its size distribution, or both.
One material parameter that influences both of these properties is the
particle size distribution (PSD) of the cement. Naturally, the size
distribution of the cement particles has a large influence on the initial and
hydrated pore size distribution of the cement paste. While the volume of
chemical shrinkage is proportional to hydration [1] and would be
expected to be independent of cement PSD at equal degrees of hydration, the
availability of external water does depend on the cement PSD in the following
manner. In measuring the chemical shrinkage of a variety of cement pastes,
Geiker [5] has shown that for a given w/c ratio,
there exists a
critical degree of hydration beyond which the rate of water ingress is unable
to keep up with that needed to maintain saturation. This point corresponds
to the depercolation of the capillary pore space [6,7]. Once
the capillary porosity becomes
disconnected, water must penetrate via the much smaller pores in the
C-S-H gel, so that the penetration rate slows by one order of magnitude
or more. Recently, simulation studies have indicated that this depercolation
of the capillary porosity is dependent on the cement PSD
[8], with coarser
cements requiring a larger degree of hydration to achieve depercolation.
Thus, under proper curing conditions, a coarser cement may result in a
reduction in both the empty porosity created by chemical shrinkage and the
autogeneous shrinkage of the specimen. In this paper,
computer simulations are applied to studying the influence of cement PSD on
capillary porosity percolation, empty porosity created by chemical shrinkage,
and interfacial transition zone (ITZ) microstructure.
Next: Computer Modelling Techniques Up: Main Previous: Main