The Alkali-Silica Reaction in Alkali-Activated Fly Ash Concrete

Open Access
Neves, Juliana Moraes
Graduate Program:
Civil Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
March 17, 2016
Committee Members:
  • Farshad Rajabipour, Thesis Advisor
  • Alkali-silica reaction
  • alkali-activated
  • fly ash
  • pore solution
The global concrete production has grown considerably over the last decades in line with the population growth, industrialization of developing countries, and the need for more infrastructures. In addition to replacing the natural environment by roads and buildings, carbon dioxide emission and depletion of natural resources for manufacturing portland cement, for example, are of major concern. The best approach to minimize the environmental impacts caused by the concrete industry is to build structures that are durable. Another valuable strategy is to manufacture concrete by using industrial by products, such as fly ash, which may fully replace portland cement. The combination of the two approaches is ideal and even more promising towards making concrete a more sustainable man-made material. This research investigates the risk of alkali-silica reaction (ASR) in alkali-activated fly ash concrete. ASR is a major deterioration mechanism, which shortens the service life of concrete structures. It involves a reaction between metastable (e.g. poorly crystalized) forms of silica in aggregates and the highly alkaline pore solution of concrete. The product of this reaction is formation of an expansive ASR gel, which cracks and damages the concrete structure. Alkali-activated fly ash (AAFA) belongs to a new generation of green concrete binders that fully replace the ordinary portland cement. AAFA binders require a highly alkaline solution to promote hydration of fly ash and strength development, which raises the concern for ASR. In this research, the concrete prism test (ASTM C1293) was used to evaluate the ASR risk of two structural grade AAFA concretes. Despite their initially high pH and presence of highly reactive aggregate, ASTM C1293 results showed absence of deleterious expansion in these two AAFA concrete mixtures (FA1 and FA2). On the other hand, the control (i.e. OPC-based) mixture, proportioned with the same amount of reactive aggregate, exceeded the expansion threshold early during the test. SEM micrographs were used to assess the extent of aggregate deterioration and ASR gel formation in the tree mixes. The SEM micrographs reveal that aggregates in FA1 concrete were more preserved than in FA2, where very little ASR gel was detected. Moreover, EDS quantitative analysis detected increased amount of alkalis in residual aggregates with concentrations similar to that found in ASR gel formed in OPC concrete. To understand the mechanism leading to absence of ASR expansion, even though there is aggregate deterioration, microstructural investigation (MIP) and pore solution analysis were performed in AAFA pastes to test four proposed hypotheses. The results suggest that pH drop and abundance of dissolved aluminum decreases the alkaline attack to the aggregates in FA1, while the insufficient calcium prevents polymerization of dissolved silica from aggregates in FA2. In comparison to OPC paste, AAFA pastes had similar or larger porosity and average pore size, despite their significantly lower ASR activity. This rules out a hypothesis that ASR is mitigated in AAFA concrete because of its low permeability. In summary, the mechanisms responsible for absence of ASR in AAFA concretes were (1) pH drop, (2) high concentration of dissolved aluminum, and (3) low concentrations of calcium in the pore solution.