Quasi-static and dynamic splitting of high-strength concretes – tensile stress–strain response and effects of strain rate

Abstract

Quasi-static and dynamic splitting tensile tests were performed on three high-strength concretes with different compressive strengths – 60 MPa, 80 MPa and 110 MPa, denoted by C60, C80 and C110 respectively – to examine their tensile properties. Quasi-static tests were conducted at a strain rate of ∼2 × 10−6 s−1, using a Shimadzu Universal Testing Machine; and dynamic tests were performed at strain rates of ∼1–10 s−1, using an 80 mm-diameter split Hopkinson pressure bar (SHPB) device. Strain gauging and digital image correlation (DIC) techniques were employed to determine the deformation in specimens. From the force applied and the average tensile strain at the specimen centre, obtained from strain gauge measurements, the uniaxial tensile stress–strain response at that location was derived, by adopting an elastic-plastic analysis, whereby tensile plastic hardening is assumed to be piecewise linear. The effects of strain rate on the tensile response were also examined and analysed. For all three high-strength concretes, significant rate dependence was observed; the yield stress, tensile strength and failure strain all increase with strain rate, and their dynamic values are much higher than their static values. The Dynamic increase factor (DIF), i.e. the ratio between the dynamic and static tensile strengths, was calculated to quantify the rate sensitivity. Similar DIFs were obtained for C60, C80 and C110, and the DIFs determined were compared with predictions based on two empirical formulae, i.e. the CEB-FIP model (2010), and that proposed by Malvar and Ross, which are frequently used for normal-strength concrete. These two formulae appear inadequate in defining the rate dependence of high-strength concrete. Consequently, a bilinear description of the DIF as a function of log10(ɛ˙), with a transition at a strain rate at 1 s−1, was proposed for high-strength concrete, based on the experimental data.