The so-called alkali-carbonate reaction (ACR) — Its mineralogical and geochemical details, with special reference to ASR

Typical examples of so-called alkali-carbonate reaction (ACR) in the Canadian field concretes in Ontario, CSA concrete prism, RILEM concrete microbars and RILEM mortar bar containing Pittsburg aggregate, were examined petrographically based on polarizing microscopy, SEM observation and quantitative SEM-EDS analysis of the reaction products. It was revealed that ASR gel was the main product responsible for the crack formation in concretes, and that this gel had a common nature to that in the typical ASR. That is, ASR gel presented distinctive compositional trend lines, passing from low-Ca ASR gel at [Ca/Si] = 1/2–1/6, [Ca]/[Na + K] = 1.0 to the “convergent point” with [Ca/Si] = 1.3–1.8, [Ca]/[Na + K] = 100 at which chemical equilibrium is attained with CSH gel. The so-called ACR is a combination of deleteriously expansive alkali-silica reaction (ASR) of cryptocrystalline quartz, and harmless dedolomitization which produces brucite and carbonate halo. In laboratory specimens, fine dolomitic aggregate undergoes dedolomitization, and brucite and ASR gel react to form non-expansive Mg-silicate gel on the dolomite crystals. This explains why the mortar bar produces smaller expansion than the concrete microbar, and why the reaction products are so minute that they escape attention by optical microscopy. As a crystalline counterpart, mountainite is a candidate for low-Ca ASR gel, while sepiolite is one for Mg-silicate gel. Concealed ASR was detected in ACR-affected field concretes undergoing ingress of deicing salt which formed Friedelʹs salt and Cl-doped CSH gel. Compositions of ASR products, methods of sample preparation and analysis for correct identification of ACR, and artifacts were critically reviewed.