Civil Engineering Reference
In-Depth Information
Pz
b
Pz
a
Mineral admixture content (%)
8.5
Schematic representation of the effect of pozzolan on ASR-
expansion (inspired from Thomas and Folliard, 2007; Thomas 2011;
Carles-Gibergues and Hornain, 2008).
The amount of SCMs required to control ASR-expansion in concrete
increases for the following conditions (from
Pz
a
to
Pz
b
on Fig. 8.5):
∑ For a given pozzolan, when the amount of alkalis in the mixture (from
cement or other internal or external sources) is high or becomes higher, the
aggregates are more alkali-reactive, or the temperature of the test decreases
(e.g. accelerated tests at 40 or 60 °C vs.
in situ
conservation).
∑ For a given mixture kept in fixed conditions, when the pozzolan is less
reactive (less amorphous silica) or contains more alkalis or calcium.
In the case of very active pozzolans, the required content is lower (silica
fume: 8-12%, metakaolin: 10-20%) than for less reactive SCMs (low-calcium
fly ash: 20-30%; high-calcium fly ash: 40-60%; GGBS: 35-65%) (Thomas
and Folliard, 2007).
Several mechanisms have been put forward to explain the effectiveness
of SCMs against ASR. It is more than probable that many of them should
be taken simultaneously to justify the role of SCMs.
∑ A densification of the hardened paste due to the physical and chemical
actions of SCMs leads to a decrease in the porosity and the permeability
(see Section 8.2.1) and the resulting reduction of the ionic mobility
slows down the migration of alkalis towards the reactive aggregate. At
the same time, these C-S-H lead to an improvement in the strength of
the cement paste and, consequently, a higher resistance to the expansive
stresses due to ASR gels (Turriziani, 1986; Bérubé and Duchesne, 1992;
Glasser, 1992).
∑ A dilution of alkalis in concrete when low-alkali SCMs are used as cement
replacement (Glasser, 1992). According to this hypothesis, pozzolans
having high alkali contents, such as some silica fumes (Na
2
O
eq
> 3%),
