Enhanced crack-bridging by unbonded inclusions in a brittle matrix

A method of engineering the microstructure of brittle composites is proposed and tested that improves the effectiveness of non-fibrous inclusions to bridge cracks in the matrix. This microstructure consists of a continuous matrix of brittle material surrounding brittle inclusions that are not bonded to it. In order to demonstrate that this works, a series of experiments have been performed with model cement systems, including aggregates such as: coated and uncoated sand; spherical and angular glass; and expanded polystyrene balls. The results have shown that the interfacial bonding between the inclusions and the matrix controls the crack path through the composite and determines whether or not an inclusion will act as a bridge and consequently provide a toughening effect. Specifically, when a moderately strong bond is present the crack path is such as to reduce the number of bridging inclusions. However, when there is no bond at all, the matrix crack follows a flatter path and more inclusions bridge the crack. Nevertheless bridging is only effective with non-spherical particles. The experimental results also suggest, that whilst the initial toughness is reduced by very weak interfaces, such composites are most likely to exhibit rising toughness behaviour. The technological implications for cementitious and refractory materials are discussed.