Using Numerical Modelling and Artificial Intelligence for Predicting the Degradation of the Resistance to Vertical Shear in Steel – Concrete Composite Beams Under Fire

Abstract

In this paper the combination of numerical modelling and artificial intelligence for predicting the degradation of the resistance to vertical shear in composite beams under fire is exposed. This work presents a technique that integrates the backpropagation learning method with a method to calculate the initial weights in order to train the Multilayer Perceptron Model (MLP). The method used to calculate the initial weights of the MLP is based on the quality of similarity measure proposed on the framework of the extended Rough Set Theory (RST). The artificial neural network models were trained and tested using numerical results from the thermal-structural analysis carried out by the two-dimensional fire-dedicated FE software Super Tempcalc and the computational tool SCBEAM which was developed by the authors for the resistances determinations. The results revealed that the proposed approach accurately permits the prediction of the degradation of the resistance to vertical shear in steel – concrete composite beams under fire.