Next: Use of cement
Up: Modifications of ITZ
Previous: Use of mineral
Recently, Zhang and Gjorv  have characterized the microstructure of the interfacial zone in lightweight concrete as being more dense and homogeneous than that found in normal weight concrete. One possible reason for this observation is the rearrangement of the cement particles in the vicinity of an aggregate due to water absorption by the aggregate. As discussed earlier, this phenomenon can be conveniently simulated using the microstructure model.
To perform this investigation, the coarse aggregate was assumed to have an absorption of 14% on a dry mass basis (data from Ben-Othman and Buenfeld ) and to be present in the concrete at a 42% volume fraction. Assuming that dry aggregate was added to the mix, an original w/c ratio of 0.586 would be required to produce a final w/c ratio in the paste of 0.39 after aggregate saturation. The final unhydrated configuration for such a system is given in Fig. 5a. Here, cement particles were placed to achieve the starting 0.586 w/c ratio and then were moved a pixel at a time towards the flat plate aggregate surface until the final w/c was produced. Of course, mixing of concrete would be expected to disturb this arrangement of particles, but this simple approach should adequately serve as a model representation of the phenomenon. This system can be contrasted against a system with normal weight aggregate and a w/c of 0.39.
Figure 5: Hydration of lightweight absorptive aggregate concrete (a) initial microstructure (top), and (b) quantitative phase analysis (bottom) (normal concrete included for comparison).
Results for phase fractions vs. distance from the aggregate surface after 77% hydration are provided in Fig. 5b for both porosity and C3S plus CSH. The lightweight absorptive aggregate improves both of these phase distributions relative to those found for the normal aggregate system. Because the cement particles are more densely packed near the aggregate, as the aggregate is approached from a distance the porosity decreases slightly while the C3S + CSH volume fraction increases slightly. Within 5 to 10 pixels of the aggregate, the porosity increases and C3S + CSH volume fraction decreases but not as much as for the system with normal weight aggregate. Thus the lightweight absorptive aggregate partially eliminates the wall effect although it is still present to an extent as even packing spheres against a flat wall results in a high porosity zone near the wall surface. Here, the scale of the roughness of the cement particles and the aggregate may play a role in real concrete. In addition to roughness providing mechanical interlocking, for absorptive aggregate, one might expect the cement particles to be drawn tightly against the rough aggregate surface in a "lock and key" fashion to substantially reduce the wall effect. In any case, the one-sided growth mechanism would still be operative.