Scaling behaviour and dual renormalization of experimental tensile softening responses

The aim of the present paper is to investigate the scaling behaviour of the experimental tensile softening (or catastrophic) curves and it concludes a series of papers from the same research group published in Materials and Structures [1-4]. Experimentally, it is evident how these curves become steeper by increasing the size of the specimens (ductile-to-brittle transition). To study this phenomenon, a completely new testing apparatus, made up of three orthogonally placed actuators, was built at the Politecnico di Torino by the authors [1, 5-7]. This set-up makes it possible to apply a purely tensile force, so that the secondary flexural stresses, if kept under control, constitute a degree of error comparable to the values allowed for normal testing apparatus. The method enables a stress vs. strain curve to be plotted with the descending (softening) branch up to the point where the cross section of the tensile specimen breaks away. The results of a new experimental investigation, performed over a very large scale-range of unnotched concrete specimens (16.1), will be presented. Particular attention will, be paid to the scale effects both on nominal tensile strength and on fracture energy. The renormalized experimental curves will be presented in a load vs, displacement plane characterized by anomalous physical dimensions. The renormalization of experimental curves enables obtaining the same response for all specimen dimensions, and then defining the universal (scale-independent) behaviour of the material.