Evaluation of the chemical index model for predicting supplementary cementitious material dosage to prevent the alkali-silica reaction in concrete

This study investigates the chemical index model effectiveness in predicting the minimum replacement of several types of Supplementary Cementitious Materials (SCMs) to mitigate the potential Alkali-Silica Reaction (ASR) in concrete. The experimental program results evaluating four natural aggregates (AF1, AF2, AF3, and AF4) and two Recycled Concrete Aggregates (RCA1 and RCA2), as presented here, were used in the model predictions. The SCMs used in this study consisted of fly ash class C (FA1), fly ash class F (FA2), ground granulated blast-furnace slag (SL), and silica fume (SF). For each SCMs, three levels of Portland cement replacements were used in the production of mortar bars: 15%-20%-25% for FA1-FA2; 20%-40%-60% for SL; and 5%-8%-10% for SF. The natural aggregates were tested according to the ASTM C1260 accelerated mortar bar test. For the RCA, an alternative preparation procedure was applied in the experimental phase to eliminate variability in the results given the RCA's high absorption. Further testing was then implemented for the aggregates classified as potentially reactive by evaluating different SCMs replacement levels to mitigate ASR using the accelerated mortar bar method, following the ASTM C1567. A Monte Carlo simulation was conducted to evaluate the chemical index prediction uncertainty, based on the model inputs statistical variation. The SCMs evaluated in this investigation were effective in mitigating ASR while used in sufficient proportion. The most effective SCMs in reducing ASR was silica fume, with a 10% replacement of Portland cement's weight, fly ash class F was next (15-25%), and finally ground granulated blast-furnace slag (40-60%). The chemical index has proven to give good results when predicting the minimum replacement of supplementary cementitious materials required to suppress ASR based on its chemical composition. (C) 2020 Elsevier Ltd. All rights reserved.