136 Damage to concrete structures
volatile and not exactly known. While Powers and Steinour (1955) gave a
fundamental explanation of the complex reaction schemes, they more or
less restricted ASR to topochemical processes. It is now accepted that ionic
reactions within the pore solution are also of great importance to ASR.
A multi-stage process is typically considered now to explain the mechanism
of expansive ASR (Bazant and Steffens 2000).
The first stage of the reaction scheme is the hydrolysis of the amorphous
silica by the hydroxyl ions, resulting in the formation of a gel-like layer on
the surface of the aggregates. For highly alkaline solutions, the monomer
H 2 SiO 4
2− is formed, while a somewhat lower pH (toward 11.2) leads to the
monomer H 3 SiO 4
− .
In the second stage, the negatively charged gel species attract positive
ions from the pore solution. Depending on the type of cations, swelling
will occur or not. In a calcium rich pore solution, Ca 2+ will be attracted
by the gel, resulting in the formation of calcium silicate hydrate, C-S-H,
thus transforming the gel into a rigid structure. This process is similar to
puzzolanic reaction schemes and will not cause swelling damage. When the
pore solution is poor in calcium, mainly the alkali ions (Na + and K + ) are
attracted by the gel. A viscous and expanding gel is produced, potentially
causing damage to the concrete.
Ichikawa and Miura (2007) further detail the mechanism involved in
swelling damage of ASR, as illustrated in Figure 5.25. Within their modified ASR model, the first stages are similar to what has been previously
mentioned: hydrolysis of the amorphous silicate and subsequent attraction
of alkali ions into the gel, forming a hydrated alkali silica gel. But in order
to cause swelling damage, further steps are required, according to Ichikawa
and Miura (2007). Due to the reaction of OH − with the amorphous silica
and the uptake of alkali ions by the gel, the solubility of Ca(OH) 2 will
increase, leading to an increased content of Ca 2+ ions in the pore solution.
As a result, Ca 2+ ions penetrate into the soft alkali silica gel, forming a rigid
rim of C-S-H around the reactive aggregate particle. This rim, however,
still allows the diffusion of alkali and hydroxyl ions to the inner soft gel,
making it further expand. Inside the aggregate, a highly expansive pressure
is now developed, and cracking will occur when the pressure exceeds the
strength of the aggregate surrounded by the reaction rim and the cement
paste.
It is further stated by Ichikawa and Miura (2007) that ASR will not
result in the deterioration of concrete when the formation of the alkali
silica gel is completed before the formation of the rigid reaction rim around
the aggregates. As a result, reactive but tiny silica-rich aggregates (e.g. fly
ash) will not induce damage because they are entirely converted to alkali
silicate before the formation of the reaction rim. With this conclusion, the
author show again the resemblance of puzzolanic reactions and the alkali
silica reaction.
volatile and not exactly known. While Powers and Steinour (1955) gave a
fundamental explanation of the complex reaction schemes, they more or
less restricted ASR to topochemical processes. It is now accepted that ionic
reactions within the pore solution are also of great importance to ASR.
A multi-stage process is typically considered now to explain the mechanism
of expansive ASR (Bazant and Steffens 2000).
The first stage of the reaction scheme is the hydrolysis of the amorphous
silica by the hydroxyl ions, resulting in the formation of a gel-like layer on
the surface of the aggregates. For highly alkaline solutions, the monomer
H 2 SiO 4
2− is formed, while a somewhat lower pH (toward 11.2) leads to the
monomer H 3 SiO 4
− .
In the second stage, the negatively charged gel species attract positive
ions from the pore solution. Depending on the type of cations, swelling
will occur or not. In a calcium rich pore solution, Ca 2+ will be attracted
by the gel, resulting in the formation of calcium silicate hydrate, C-S-H,
thus transforming the gel into a rigid structure. This process is similar to
puzzolanic reaction schemes and will not cause swelling damage. When the
pore solution is poor in calcium, mainly the alkali ions (Na + and K + ) are
attracted by the gel. A viscous and expanding gel is produced, potentially
causing damage to the concrete.
Ichikawa and Miura (2007) further detail the mechanism involved in
swelling damage of ASR, as illustrated in Figure 5.25. Within their modified ASR model, the first stages are similar to what has been previously
mentioned: hydrolysis of the amorphous silicate and subsequent attraction
of alkali ions into the gel, forming a hydrated alkali silica gel. But in order
to cause swelling damage, further steps are required, according to Ichikawa
and Miura (2007). Due to the reaction of OH − with the amorphous silica
and the uptake of alkali ions by the gel, the solubility of Ca(OH) 2 will
increase, leading to an increased content of Ca 2+ ions in the pore solution.
As a result, Ca 2+ ions penetrate into the soft alkali silica gel, forming a rigid
rim of C-S-H around the reactive aggregate particle. This rim, however,
still allows the diffusion of alkali and hydroxyl ions to the inner soft gel,
making it further expand. Inside the aggregate, a highly expansive pressure
is now developed, and cracking will occur when the pressure exceeds the
strength of the aggregate surrounded by the reaction rim and the cement
paste.
It is further stated by Ichikawa and Miura (2007) that ASR will not
result in the deterioration of concrete when the formation of the alkali
silica gel is completed before the formation of the rigid reaction rim around
the aggregates. As a result, reactive but tiny silica-rich aggregates (e.g. fly
ash) will not induce damage because they are entirely converted to alkali
silicate before the formation of the reaction rim. With this conclusion, the
author show again the resemblance of puzzolanic reactions and the alkali
silica reaction.