Author(s)

I. Penis, F. Casciati & M. Battaini

Abstract

Modelling of uncertainty in the behaviour of RC structural elements I. Penis," F. Casciati,* M. Battaini* o Department of Civil Engineering, University of Ljubljana, 61001 Ljubljana, Slovenia * Department of Structural Mechanics, University of Pavia, 127100 Pavia, Italy ABSTRACT A large databases of results of laboratory tests on reinforced concrete experi- ments is gathered. Attention is focused on ultimate rotation, i.e. ductility, and low-cycle fatigue capacities. These are regarded as the result of a mapping from the space of the input variables which identify the single test. The standard regressive approach to the definition of the mapping is replaced by artificial in- telligence techniques, as neural networks and neural network-like systems. INTRODUCTION The deformation capacity of the structural elements has been widely recognized as the main source of damage for a R.C. frame (e.g. Bertero et al., 1991; Qi and Moehle, 1991). An appropriate (probabilisti

Keywords