This study had two objectives: (1) to evaluate the effectiveness of particular mineral admixtures when combined with portland cements of varying alkali content in preventing expansion due to alkali-silica reactivity (ASR), and (2) to determine if set minimum amounts of various types of mineral admixtures could be established for use with cements of varying alkali content to provide protection from ASR. Mortar bars made with Pyrex glass aggregate were stored moist at 38 ° C for 366 days and measured periodically for length change. Control mortars were mixed with portland cements having an alkali content ranging from 0.18% to 0.92% Na20 equivalent. Test mortars were mixed replacing various percentages of portland cement with Class F fly ash (15%, 25%, and 35%), slag (20%, 35%, and 50%), or silica fume (3% and 7%). Test mortars were also mixed using IP cement produced using Class F fly ash (18%, 20%, and 22%). Expansions of control mortars with alkali contents of 0.4% or less were negligible through the 336-day test period. With cements having an alkali content above 0.4%, equilibrium between cement alkali content and mortar bar expansion was reached at 56 days, when a strong linear relationship between increasing expansion and increasing alkali content developed. A regression analysis was performed on data from the control batches, and the resulting equation was used to normalize data from the test batches. The expansion of mortars containing fly ash, slag, and silica fume was lower than the expansion of the control batches for portland cement having an alkali content of 0.6% and above for all replacement levels of the tested materials. The amount of expansion varied with the type of material, percentage replacement, and cement alkali content. Using an expansion of 0.10% as the maximum indicative of acceptable performance in inhibiting ASR, minimum amounts of Class F fly ash, slag, and silica fume were established for use with portland cement depending on the alkali content of the cement.

The effects of silica fume and fly ash on reducing the risk of concrete damage due to alkali-silica reaction are presented in this thesis. One silica fume and twenty fly ashes of a wide range of chemical composition were investigated. The effect of the selected materials on expansion when used with high-alkali Portland cement and a siliceous limestone reactive aggregate (Spratt) was evaluated using the concrete prism and the accelerated mortar bar tests. Portland cement pastes and lime/pozzolan blends were prepared using similar combinations of materials used in the expansion tests. These samples were used to study the effects of fly ash and silica fume on the pore solution alkalinity and composition of hydration products. The study showed that the efficiency of fly ash in limiting the risk of deleterious expansion is linked to its efficiency in lowering the pore solution alkalinity. High-calcium and high-alkali fly ashes are generally less effective in this respect compared to low-calcium ash. Silica fume was found to be effective in reducing the pore solution alkalinity at early ages; however, the alkalinity was found to increase at ages later than 28 days. Combinations of moderate levels of silica fume (5%) and high-calcium fly ash (20-30%) were found to be effective in limiting the expansion even after three years. This was attributable to the high capacity of such combinations to achieve and maintain low pore solution alkalinity. The silica fume reduces the pore solution alkalinity at early ages and the fly ash maintained the low alkalinity at later ages. The study showed that, for low and moderate-alkali fly ash, the 14-day expansion in the accelerated mortar bar test could be used to predict the 2-year expansion of concrete prisms. The results from the lime/pozzolan blends confirmed that the hydration products of pozzolan play the dominant role in reducing the pore solution alkalinity of cementitious systems containing pozzolans. The two-year expansion of concrete prisms containing fly ash could be correlated reliably to the alkalis available from the ash as determined by leaching mature paste samples in leaching solution of alkali ion concentration of 0.25 moles per litre.