The work reported in this paper is part of an ongoing program to investigate ASR at a fundamental level, by measuring the generated stress rather than just the sample expansion. However, the maximum stress that can be measured with the load cell available to us is about 8 MPa, for this size specimen. Therefore, we cannot determine if the stress will reach an equilibrium value for the samples without silica fume. We are planning to install a higher capacity load cell to further investigate this behavior. More SEM observations will be made to study the crack patterns and matrix composition. In addition, we will also study, using a combination of experimental and modelling techniques, the fundamental mechanisms involved in ASR deterioration.
Silica fume, because of its amorphous nature and large surface area, reacts rapidly with the alkalies in the pore solution, forming C-S-H by reacting with the calcium and hydroxide ions. This is essentially ASR. Why does this reaction not cause expansion? With aggregates, the reaction is slower due to the reduced surface area per volume, and more concentrated, thus promoting the build up of higher stresses and consequent cracking. Therefore, the particle size distribution of the aggregates should be investigated.
While the addition of silica fume as a replacement of cement reduces the stresses and the expansion generated by the ASR reaction, expansion and stress are not completely eliminated, at least at the replacement dosage used of 15 % by mass (a dosage that is relatively high with respect to common practice). Due to continuing reaction of the aggregates in the specimens containing silica fume, the number of cracks is reduced but not eliminated. Future work is planned to determine the fundamental mechanisms of silica fume mitigation of the ASR and the potential effect of the residual expansion and cracking.