Recently, alkali-activated cement (AAC) has been studied to partially replace portland cement (PC) to reduce the environmental impact caused by civil construction and the cement industry. However, with regard to durability, few studies have addressed the behavior of AAC. This study aimed to evaluate the performance of AAC made from blast-furnace slag with contents of 4 and 5% sodium hydroxide as an activator (Na2Oeq of 3.72% and 4.42%, respectively) when subjected to alkali-aggregate reaction (AAR). Length variation tests were carried out on mortar bars immersed in NaOH solution (1 N of NaOH, T = 80°C [176°F]) and on concrete bars (T = 60°C [140°F], RH = 95%); compressive strengths tests and scanning electron microscopy-energy dispersive X-ray spectroscopy (SEM-EDS) analyses were also made. Two types of PC were used as a comparison. The results showed good behavior of the AAC in relation to the AAR, with expansions lower than those established by the norm (34% of the limit) and without the finding of losses of mechanical resistance or structural integrity. The alkaline activator content had a small influence on the behavior of the AACs, in which the lowest amount of NaOH (4%) showed fewer expansions (only 15% of the limit established by the norm). Even for the highest activator content (5%), the results were good and comparable to those of PC with pozzolans, which is recommended for the inhibition of AAR.

The mechanism of alkali-carbonate reaction (ACR) is still controversial. ACR distress in concrete has been described as an increase in volume caused by the crystallization of brucite following dedolomitization. In this study, the cause of concrete distress in reported ACR-damaged concrete pavements was investigated, and it was determined that alkali-silica reaction (ASR) was the cause of the damage. Optical microscopy and scanning electron microscopy with energy-dispersive spectroscopy (SEM-EDS) analyses identified ASR gel extending from reactive aggregates into the paste, X-ray elemental mapping confirmed the composition of the gel, and EDS determined the amount of each element in the ASR gel spectra. Silica in the form of cryptocrystalline-microcrystalline quartz was found in the matrix of reactive aggregates and was the source of reactive silica. The test results confirmed that ASR caused the damage to the primary concrete pavements and present the first case ever reported in the United States in which ASR is the main cause of concrete damage in concrete made from carbonate aggregate exhibiting a classic texture and composition cited for ACR.

This paper presents tests that can be used collectively to provide a qualitative assessment of residual expansion in structures affected by alkali-silica reaction (ASR). The tests are applied to bridge barriers suffering different levels of ASR deterioration. These include testing extracted cores under different lab conditions, monitoring concrete elements under field condition, damage rating index (DRI) on cores, and measuring alkali levels in the affected concrete. Expansion of barriers with low deterioration level was double that of highly deteriorated barriers at 4.5 years. Similar results were reached through testing cores under laboratory conditions at 38°C (100°F) and 100% relative humidity, although the DRI showed the same increase in damage in both cores after testing. Testing cores under laboratory conditions until expansion ceases helps in predicting the minimum residual expansion. Soaking cores in alkaline solutions of different concentrations and finding the level required to trigger expansion helps in assessing the risk of future expansion.

The reference Pittsburg Ontario alkali carbonate-reactive (ACR) aggregate source was characterized using a holistic approach to identify and quantify mineral phases, particularly reactive forms of silica and expansive types of clays in the aggregate which may have a role in the controversial ACR mechanism and resulting expansion and cracking of concrete. This research was performed using state-of-the-art analytical techniques and methods that included polarized light microscopy (PLM), quantitative image analysis (QIA) of backscattered electron (BSE) scanning electron microscope (SEM) images, acid-insoluble residue (AIR) tests, X-ray diffraction (XRD), X-ray fluorescence (XRF), and a chemical ratio method of identifying alkali-reactive carbonate rocks. Three types of silica phases were identified through PLM examination: upper silt-sized quartz grains both in the virgin aggregate and acid insoluble residue (AIR); cryptocrystalline silica dispersed and hidden in the fine-grained rock matrix and identified only in the AIR; and cryptocrystalline-to-microcrystalline silica occupying interstitial spaces of dolomitic limestone particles which lacked clay in their matrix. PLM findings were confirmed through QIA of the AIR. Particle size distribution of silica phases through QIA showed that silica phases in sizes of 0.5 to 2 μm (0.00002 to 0.00008 in.) occur in high abundance. QIA of AIR identified illite as the major clay mineral in the aggregate. While this clay type is not known to be expansive, microcrystalline to cryptocrystalline silica phases are potentially alkali-silica reactive (ASR) in concrete as opposed to ACR.

Alkali-silica reaction (ASR) is one of the most harmful distress mechanisms affecting concrete infrastructure worldwide. ASR is a chemical reaction that generates a secondary product, which induces expansive pressure within the reacting aggregate material and adjacent cement paste upon moisture uptake, leading to cracking, loss of material integrity, and functionality of the affected structure. In this work, a computational homogenization approach is proposed to model the impact of ASR-induced cracking on concrete stiffness as a function of its development. A representative volume element (RVE) of the material at the mesoscale is developed, which enables the input of the cracking pattern and extent observed from a series of experimental testing. The model is appraised on concrete mixtures presenting different mechanical properties and incorporating reactive coarse aggregates. The results have been compared with experimental results reported in the literature. The case studies considered for the analysis show that stiffness reduction of ASR-affected concrete presenting distinct damage degrees can be captured using the proposed mesoscale model as the predictions of the proposed methodology fall in between the upper and lower bounds of the experimental results.