Size effect in splitting diagonal cubes

Abstract

The experimental investigations on fracture mechanics of cement-based materials until 1970s indicated that classical linear elastic fracture mechanics (LEFM) is invalid for quasibrittle materials such as concrete. This inapplicability of LEFM is due to existence of an inelastic zone with large scale and full cracks in front of the main crack tip in concrete. This so-called fracture process zone (FPZ) is ignored by LEFM. Consequently, several investigators have developed non-linear fracture mechanics approaches to describe failure of concrete/reinforced concrete structures. Deterministic size effect laws among these non-linear approaches, for instance size effect law (SEL) by Bazant (1984), suggest that size effect on strength is primarily related to a relatively large FPZ in concrete. One of the main requirements in this law is the need to test samples, which are geometrically similar and made of the same material, and which must provide a minimum size range=1:4. The split-tension test has been used to indirectly test the tensile strength of quasi-brittle materials such as concrete and rock. Recently, concrete splitting specimens have been commonly used in concrete fracture because they have certain advantages, such as compactness and lightness, compared to beams. However, the number of theoretical and experimental studies with diagonal split-tension specimens, to which compressive forces are applied along two opposite edges, is limited. In this study, two series of concrete diagonal cube specimens of different size (size range 1:4) were tested by splitting loading. The concrete mixes with the maximum aggregate size=8 mm were designed as the gap-graded aggregate and the continuously graded aggregate. The ultimate loads obtained from the test results were analysed via Bazant’s SEL. Consequently, it was observed from the analysis based on SEL that the concrete with gapgraded aggregate is the more ductile material than the concrete with continuously graded aggregate.