Report of RILEM TC 301-ASR: Modelling the impact of SCMs, alkali level and w/b ratio on alkali concentration in pore solution

Abstract

Supplementary cementitious materials (SCMs) can mitigate alkali-silica reaction by lowering the alkali metal concentration in the pore solution. This is a theoretical study on the applicability of a thermodynamic model (GEMS) and the empirical Taylor model to predict the required replacement level of portland cement (PC) by SCMs to achieve an alkali metal concentration below 300 mmol/L. The SCMs investigated are silica fume (SF), metakaolin (MK), fly ash (FA) and slag. The impact of the alkali content of the PC and the w/b ratio on the required replacement level is modelled and compared to experimental pore solution concentrations. Both models predict a similar impact of the SCM replacement level on the distribution of alkali between the pore solution, C–S–H and unreacted material. The thermodynamic model predicts little impact of the alkali content of PC and the w/b-ratio on the required replacement level, i.e., 20% SF, 20% MK, 40–50% FA and 60–70% slag. This is contrary to the Taylor model, which predicts that the replacement levels of FA and slag ranges from 7 to 58% when increasing the alkali content from 0.47 to 0.93% and from 80 to 10%, when increasing the w/b ratio from 0.3 to 0.9. The required replacement levels for SF and MK vary between 2 and 19% when increasing the alkali content from 0.47 to 0.93%, and from 40 to < 5% when increasing the w/b ratio from 0.3 to 0.9. The main difference between the two models is how they account for the uptake of alkali metals by the C–S–H.