Figure 2: Porosity vs. distance from aggregate surface before and after hydration for (left) wall effect and one-sided growth mechanism, and (right) one-sided growth mechanism only.
Next: Modifications of ITZ
Up: Main
Previous: Model description
The application of the microstructure model to simulate ITZ's has been described by Garboczi and Bentz [12]. The results of a simulation similar to the system in Fig. 1, but with circular cement particles, indicated that the ITZ's contain less cement and CSH and more porosity and CH than the bulk paste, in accord with what has been observed in real concretes. Although particle packing effects obviously contributed to these characteristic features of the ITZ's, a secondary one-sided growth mechanism also plays a role. In the interfacial transition zones, available porosity is filled with hydration products growing from one direction only, unlike the bulk paste where products grow inward from all directions. Using simulation, it was possible to separate these two effects. By first placing the cement particles throughout the entire system and then overlaying the aggregate and eliminating all portions of cement particles underneath the aggregate, it was possible to remove the effects of the initial particle packing (i.e. the wall effect) and study only the one-sided growth effect. This system could then be contrasted against one in which the aggregate was placed first followed by the cement particles, where both the wall effect and one-sided growth mechanism would be present.
Figure 2 shows the results of this simulation in terms of porosity as a function of distance from the aggregate surface for the initial and fully hydrated systems for both system configurations mentioned above. The one-sided growth mechanism appears to be a secondary but significant contributor to the increased porosity in the ITZ. The cement particles used in this simulation were monosize and 21 pixels in diameter, so that the thickness of the ITZ was about the same as the diameter of the cement particles.
Figure 2: Porosity vs. distance from aggregate surface before and after hydration for (left) wall effect and one-sided growth mechanism, and (right) one-sided growth mechanism only.
To further investigate the link between cement particle size and ITZ thickness, a continuum 3D spherical particle parking program was implemented. Spherical cement particles of a known size distribution (from a sieve analysis) were placed randomly in order of largest to smallest in a cube 200 micrometers on a side, such that no particles overlapped. Rigid walls were present on two opposite faces of the box to represent aggregates while periodic boundary conditions are maintained on the other four sides. Volume fractions of cement and porosity were assessed using point sampling techniques. Using sieve size distributions presented in van Breugel [13], results are shown in Figure 3 for two cements at various w/c ratios. Cement A1 has a median particle size, on a mass basis, of about 28 micrometers, while the median particle size of cement A7 is about 11 micrometers. The ITZ thicknesses (i.e. the distance at which the porosity first approaches its bulk value) correspond closely to these median particle diameters. Conversely, w/c ratio has little effect on ITZ theickness, although it does affect the gradient in cement content as the aggregate surface is approached. Although hydration may decrease these ITZ thicknesses as seen in Fig. 2, these results suggest cement fineness to be one variable that strongly influences ITZ characteristics as the ITZ thickness is expected to be proportional to the median cement particle size.
Figure 3: Porosity near aggregate before any hydration, for two different cements.
Having outlined two causes of ITZ microstructural features, namely the wall and one-sided growth effects, we now explore various methods for improving ITZ microstructure.