Influence of contact friction in compression tests of concrete samples

Abstract

This paper presents an experimental assessment of the influence of the boundary conditions and geometry of the specimen in unconfined compression tests, employing two concrete types. For such purposes, an experimental program was designed, dividing the samples into cylinders with height h = 20 cm and diameter b = 10 cm, prisms with height h = 20 cm and square 10 cm base and 10 cm sides cubes. Concrete traces resulting in 50 MPa (C50) and 30 MPa (C30) nominal strengths were used. The compression tests were carried out under five different boundary conditions on the loaded planes: Grease, Teflon and Brush plate, as anti-friction means. For reference, standard tests and increasing the lateral restraint employing Glued sheets, in order to maximize the interface friction, were also performed. The strength difference between height/base ratios (h/b) = 2 and (h/b) = 1 prisms varies on average from 15% for C50 and 35% for C30 concretes. No significant difference was found between cylindrical and prismatic specimens with the same h/b ratio. In specimens provided with anti-friction devices, the strength variation for the different geometries and h/b ratios was almost totally eliminated, as well as the post-peak softening of the tests. The reported results prove that the post-peak softening is not a characteristic of the material but a consequence of the interface friction in the tests. Cubic specimens provided with grease to eliminate friction led to the same compressive strength than cylinders or prisms with the same cross-sectional area and therefore offer the unquestionable advantage of allowing the assessment of the influence of the orientation of casting.