Supplementary cementing materials (SCMs) are materials that contribute to the properties of concrete through either pozzolanic reaction (e.g., low-calcium fly ash, silica fume, or calcined clay), or hydraulic reaction (e.g., ground granulated blastfurnace slag, hereinafter called slag), or both (high-calcium fly ash). These materials are generally used to partially replace the Portland cement component, the level of replacement varying widely from less than 10% to more than 50% depending on the nature of the SCM.
There is a substantial body of published literature dealing with the effect of SCMs on ASR, but there is still conflicting evidence regarding the conditions under which SCMs are effective and how much is required to suppress deleterious expansion in concrete. A general statement that can be made is that nearly all SCMs can be used to prevent expansion provided they are used in sufficient quantity; the amount required varies widely depending on, among other things, the following: · the nature of the SCM (especially mineralogical and chemical composition) · the nature of the reactive aggregate (generally, the more reactive the aggregate, the higher the level of SCM required) · the availability of alkali within the concrete (i.e., from the Portland cement and other sources) · the exposure conditions of the concrete (i.e., concrete exposed to external sources of alkali may require higher levels of SCM). For example, the amount of SCM required to control expansion in a concrete with a highly reactive aggregate and high-alkali cement might vary as shown in Table 7.3.
Generally the amount of SCM required decreases as the reactive silica content of the SCM increases or as the calcium or alkali content of the SCM decreases (Fig. 7.5). In other words, an SCM with high silica and low amounts of calcium and alkali, such as silica fume, tends to be effective at low levels of replacement (Thomas and Bleszynski, 2001). Slag, on the other hand, is much less efficient due to its lower silica and higher calcium contents, and has to be used at much higher levels of replacement to control expansion (Thomas and Innis, 1998).
SCMs control expansion due to ASR by reducing the availability and, to a lesser extent, the mobility of alkalis in the concrete (Shehata et al., 1999). The alkalis, sodium and potassium, in concrete are partitioned between the solid phases (i.e., bound by the hydrates) and the liquid phases (i.e., pore solution) of the concrete. Only the alkali hydroxides in the pore solution can attack the reactive components of the concrete. The binding capacity of the hydrates is to a large extent a function of the calcium-silica ratio of the C-S-H that forms. C-S-H with a lower Ca/Si ratio has an increased alkali-binding capacity (Bhatty and Greening, 1978) since the surface charge becomes less positive as the calcium content decreases and this attracts more positive cations (Na+ and K+) from the surrounding pore solution (Glasser and Marr, 1985).
The effect of SCMs on the pore solution composition will depend on the amount of alkali (Na2Oe) they contribute, and the degree to which the presence of the material changes the Ca/Si ratio of the hydrates that form. Thus SCMs with low levels of Na2Oe and CaO, and high levels of reactive SiO2, will contribute little alkali to the system, but significantly increase the alkali-binding capacity of the hydrates (by reducing the Ca/Si ratio). Figure 7.6 shows the hydroxyl ion content of the pore solution extracted from 2-year-old hydrated cement paste samples that were produced with w/cm ˆ 0.50 and 79 different combinations of cementing materials including Portland cements of different alkali level, binary blends of Portland cement with a range of different fly ashes, slag, silica fume and natural pozzolans, and ternary blends of Portland cement with either silica fume plus fly ash or silica fume plus slag (Thomas and Shehata, 2004). It is evident that the hydroxyl (and alkali) ion concentration is a function of the alkali, calcium and silica content of the blended cement.
Concrete prism tests can be used to evaluate the efficacy of different SCMs and to determine the level of SCM required to control expansion with a specific reactive aggregate. It is generally considered necessary to extend the duration of the concrete prism test from one year (used to evaluate aggregate reactivity) to two years when evaluating SCM to ensure that expansion is prevented and not merely retarded. Concrete prism tests are generally not suitable for assessing the combination of low-alkali cement plus SCM for the reasons discussed in Section 7.7.2. Additional alkalis are required in the concrete prism test to accelerate the reaction and compensate for the loss of alkalis due to leaching during testing.
The (ultra) accelerated mortar bar test has also been used to evaluate SCM/ reactive aggregate combinations and to determine the amount of SCM required to control expansion with a particular reactive aggregate. Generally, combina- tions of reactive aggregate and SCM that expand less than 0.10% after 14 days immersion in 1M NaOH solution at 80ëC are considered to have a low risk of expansion when used in concrete under normal field conditions (Thomas and Innis, 1999).
