Avoiding & Mitigating Alkali-Aggregate Reaction (AAR) in Concrete Structures

Abstract

Alkali-Aggregate Reaction (AAR) is one of the most harmful distress mechanisms affecting the serviceability and durability of concrete critical infrastructure worldwide. Over the past decades, several approaches and recommendations have been developed to assess the potential reactivity of aggregates in the laboratory and the efficiency of preventive measures (e.g., supplementary cementing materials – SCMs) to mitigate ASR in the field. Yet, recent findings suggest that the appropriate use of SCMs “only” delayed and does not entirely prevent ASR occurrence. Moreover, once ASR starts in the field, there is no “universal” solution that should be applied in various cases, and each situation should be evaluated as “unique”. Nevertheless, artificially triggering healing agents have been studied in the late years, thus presenting an interesting “physical” solution to reduce the ingress of water and recover damaged concrete elements, which could present an interesting solution for durability-related distress due to ASR. This Ph.D. project focuses on detailed laboratory investigations aiming first to understand the self-healing process of concrete (i.e., by the natural or engineered process). Then, its further influence on ASR-induced expansion and deterioration, either applied internally or externally to the concrete. To achieve this goal, concrete mixtures presenting a wide range of binder compositions, using distinct types of chemical admixtures (e.g., crystalline self-healing), and incorporating five different types/nature of highly reactive aggregates (i.e., coarse and fine) were combined to manufactured concrete specimens in the laboratory. Otherwise, in aging specimens, concrete samples were designed only with GU-cement as the binder material but incorporated two different types/nature of highly reactive aggregates. Then, the samples were exposed to ASR-induced development until they reached pre-determined expansion levels, in which a wide range of sealers and coating materials were applied on the surface of the affected specimens. Mechanical (i.e., stiffness damage test, modulus of elasticity, micro indentation, shear and compressive strengths) and microscopic (damage rating index and scanning electron microscopy) tests were performed on samples at different ages (up to two years of accelerated ASR development). The results show that besides changing AAR-kinetics, the different binder compositions or the chemical admixtures could modify the distress mechanism due to AAR. The addition of crystalline healing agents or their combination with SCMs in concrete not only delayed the development of inner damage but significantly lowered the compressive strength loss at equivalent expansion amplitudes than control specimens. Moreover, the combination of different binder materials modified the chemical and mechanical properties of the ASR-gel, changing its swelling properties and the further damage development in concrete. On the other hand, the wide range of surface treatments used were not able to alter ASR distress mechanism; yet, they changed ASR-kinetics. Moreover, their effectiveness to slower the reaction shows to be significantly influenced by the damage degree to which the surface treatment is applied. Finally, a comprehensive framework enabling the optimized selection of raw materials to prevent or mitigate ASR development is proposed.