Abstract
The elastic modulus of concrete is utilized in the design of reinforced concrete structures, including in predicting creep deformation. This elastic modulus can be estimated, using models contained in national design codes, by considering one or more properties (usually compressive strength).
The proposed paper assesses the accuracy of eleven empirical elastic modulus estimation models, when compared with the actual values measured on a range of concretes under laboratory controlled conditions. The equations considered are those contained in BS 8110 (1985), SANS 10100 (2000), SANS 10100 (2000) Modified, ACI 209 (1992/2008), AS 3600 (1988, 2001 and 2009), CEB-FIP (1970, 1978 and 1990), EC 2 (2004), GL 2000 and 2004), GZ (1993) and RILEM Model B3 (1995).
The test results indicated that the discrepancies between the measured and estimated values were only significant in the case of the SANS 10100 (2000) Modified method (P = 3,1 %) and the CEB-FIP (1970) method (P = 2 %). The most accurate methods were the SANS 10100 (2000) and AS (2009) which both yielded a coefficient of variation (ωj) of 9,3 %. The least accurate method was the CEB-FIP (1970) which yielded a coefficient of variation (ωj) of 22,7 %.
Furthermore, the test results of this research were used to establish which factors influence the elastic modulus of concrete. It was found that the concrete density, the density of the included aggregate and the coarse aggregate content separately correlated significantly with the elastic modulus (P ≤ 3 %).