Resumen

FELIX, E. F.; CARRAZEDO, R.  y  POSSAN, E.. Parametric analysis of carbonation process in reinforced concrete structures through Artificial Neural Networks. Rev. ALCONPAT [online]. 2017, vol.7, n.3, pp.302-316. ISSN 2007-6835.  https://doi.org/10.21041/ra.v7i3.245.

The aim of this paper is parametrically analyze the main factors that influence on the progress of concrete carbonation front. Therefore, a numerical model was developed using Artificial Neural Networks (ANNs), considering the Multi-Layer Perceptron class, designed in a C++ object-oriented program. The software was fed by experimental degradation data available in the current literature. The results obtained in the parametric analysis, besides adding knowledge to the building pathology area, reinforce concepts already known in the literature, demonstrating the efficiency of ANNs in the investigation of concrete carbonation.

Palabras llave : carbonation of concrete; time-to-corrosion initiation; Artificial Neural Network; mathematical modelling..