Shape-Effect in the Effective Laws of Plain and Rubberized Concrete

The procedure of the effective law outlined in this paper [Ferretti (2001); Ferretti and Di Leo (2003); Ferretti (2004b)] is an experimental procedure for identifying the constitutive law in uniaxial compression of brittle heterogeneous materials, and is based on the physical, analytical and numerical discussions about the existence or otherwise of strain-softening [Ferretti (2004a); Ferretti (2005)]. This procedure allows us to correct several incongruities that characterize the average stress versus average strain diagrams: it produces evidence against strain-softening in uniaxial compression [Ferretti (2004b)], whose existence may be questioned from a physical point of view [Ferretti (2004a); Ferretti (2005)], it provides effective stress versus effective strain laws that are size-effect insensitive [Ferretti (2004b)] and identifies Poisson's ratio and volumetric strain, which are independent of the degree of damage during the compression test [Ferretti (2004c)], as should be the case for all constitutive parameters. The procedure also allows us to explain the gradual change of shape in the average stress versus average strain laws when a confinement pressure is applied to the specimen [Ferretti and Di Leo (2003)]. Moreover, the procedure emphasizes how the final stage in compressed concrete specimens is largely characterized by the propagation of a macro-crack, rather than by crushing. This puts a question mark on the existence of creep, which, according to the identified effective parameters, seems mainly to be a structural effect due to crack propagation [Ferretti and Di Leo (2008)]. In this paper, the identification procedure of the effective law is applied to cubic and cylindrical concrete specimens, in order to verify whether or not the effective law is sensitive to shape-effect. Two different concrete mixtures were used, the one of plain and the other of rubberized concrete. New relationships were also proposed for design purposes, both for plain and rubberized concrete.     

Boundary effect on weight function in nonlocal damage model

Some insights on boundary effects in nonlocal damage modelling are addressed. Interaction stresses that are at the origin of nonlocality are expected to vanish at the boundary of a solid, in the normal direction to this boundary. Existing models do not account for such an effect. We introduce tentative modifications of the classical nonlocal damage model aimed at accounting for this boundary layer effect in a continuum modelling setting. Computations show that some nonnegligible differences may be observed between the classical and modified formulations. In a one dimensional spalling test, only the modified formulation provides a spall of finite nonzero thickness, whereas spalls smaller than the internal length cannot be obtained according to the original formulation. For the same set of model parameters, including the internal length, the fracture energy derived from the size effect test method is also very different according to both approaches. Parameters in the size effect laws for notched and unnotched specimens, obtained from computation of geometrically similar bending beams, are more consistent with the modified nonlocal model compared to the original nonlocal formulation.     

The Cell Method: an Enriched Description of Physics Starting from the Algebraic Formulation

In several recent papers studying the Cell Method (CM), which is a numerical method based on a truly algebraic formulation, it has been shown that numerical modeling in physics can be achieved even without starting from differential equations, by using a direct algebraic formulation. In the present paper, our focus will be above all on highlighting some of the theoretical features of this algebraic formulation to show that the CM is not simply a new numerical method among many others, but a powerful numerical instrument that can be used to avoid spurious solutions in computational physics.