Next: Model Description
As concrete finds usage in new and varied applications, understanding the relationship between microstructure and properties becomes more critical. In high performance concretes, examples of which are systems based on a low water:cement (w/c) ratio and containing silica fume and a superplasticizer, the interfacial bond formed between the cement paste matrix and the small and large aggregates plays a major role in determining the ultimate performance of the composite [1,2]. Controlling the interfacial zone microstructure is thus paramount to engineering the proper performance of these new materials.
Interfacial zone microstructure has been studied experimentally by numerous researchers [3,6] and has also recently been explored using a microstructural modelling (simulation) approach [7,8]. In ordinary portland cement concrete, the interfacial zone consists of a region up to 50 micrometers thick surrounding an aggregate where the microstructure differs significantly from that of the bulk paste. In this zone, the microstructure is characterized by high porosity (consisting of relatively large pores), low cement and C-S-H (gel) contents, and large crystals of CH. This microstructural development is due to both the inefficient packing of cement particles near the aggregate surface [6,7] and the "one-sided growth" effect where, near the aggregate, hydration products are forming from one direction only in contrast to the bulk paste, where hydration products are growing inward from all directions at any given point . Other effects, such as bleeding, will only intensify this dissimilarity of microstructure of the interfacial zone relative to that of the bulk paste. This paper does not address bleeding.
In portland cement concrete containing mineral admixtures such as silica fume or fly ash, the interfacial zone microstructure has been observed to be more similar to the bulk paste microstructure than in the equivalent concrete without such admixtures [9,10]. Silica fume, for example, is particularly proficient in promoting this increased homogeneity of microstructure due to two properties: its small size (relative to the cement particles), and its high reactivity. The very small (les than 1 micrometer) silica fume particles are able to pack more closely against the aggregate surface, reducing the width of the higher porosity region that exists because of packing constraints . The high reactivity of silica fume in the subsequent pozzolanic reaction with CH produces C-S-H, which occupies more volume than the separate CH and unreacted silica fume alone. Thus, the increasing porosity gradient in the interfacial zone, although not totally removed, is reduced substantially. Additionally, the total of the unreacted cement and C-S-H phases is nearly constant throughout the system, with phase fractions in the interfacial zone only slightly less than those found in the bulk paste. This increased homogeneity of the composite in the interfacial zone should manifest itself in increased strength since a weak link no longer exists in the interfacial zone region, contributing to better stress transfer between aggregate and paste. Fly ash has similar effects, but due to its larger size and lower SiO2 content, is found to be less beneficial than silica fume in terms of improving interfacial zone microstructure .
In this paper, model and experimental results concerning the interfacial zone microstructure in portland cement concrete with and without silica fume are compared. Qualitative comparisons are made based on two-dimensional images (model and SEM) while quantitative comparisons are based on analysis of the various phase fractions as a function of distance from the aggregate surface.