Assessment of methods for optimising ternary blended concrete containing metakaolin

Abstract

The relative effectiveness of three approaches—isothermal calorimetry, trial batching and particle packing models—for optimising mix proportions in Portland cement/metakaolin (MK)/fly ash (FA) ternary blended concretes was compared through measurements of workability and compressive strength, in consideration of economy. Mixtures were prepared using single cement and class C FA and two MKs of varying fineness. The finer MK showed a higher rate of reaction, but the isothermal calorimetry also suggested that MK replacements of 5% or higher are needed to overcome the hydration delay resulting from the use of 25% by mass FA. Trial batches at three water-to-cementitious materials ratios (w/cm)—0·30, 0·40 and 0·50—were designed based on the calorimetry results. While workability decreased with increasing MK content, similar or better workability than the control mixes was achieved in the ternary blends. The compressive strength significantly increased with increasing MK content, as compared to the ordinary concrete and the FA binary blends, and especially at low w/cm (i.e. w/cm = 0·30). The trial batch strength results confirmed that the hydration rates (measured by isothermal calorimetry) could be used to optimise binder composition for early concrete strength. However, in the third approach, the particle packing model was found to be insufficiently related to measured 1-day strength or to strength in higher w/cm concrete; the closest relationship between the primary model parameter (q) and strength was found at lower w/cm and at later ages (i.e. 28 days). It is proposed that the lack of consideration of the constituent's chemical reactivity of highly reactive materials, such as MK, makes this approach unreliable at early ages and at higher w/cm. Finally, despite the lack of correlation between q and strength, results show q values of 0·26-0·29 are optimal for workability.