Bending Moment Prediction for Continuous Composite Beams by Neural Networks

Abstract

A methodology has been developed for the continuous composite beams to predict the inelastic bending moments (considering the cracking of concrete) from the elastic moments (neglecting the cracking) by using the neural networks. The proposed neural network models predict the inelastic moment ratios (ratio of inelastic moment to elastic moment) at the supports of a span. Nine significant structural parameters have been identified governing the inelastic moment ratios. Six neural networks have been presented to cover the entire practical range of the beams. The proposed neural networks have been validated for a number of beams with different number of spans and the errors are shown to be small for practical purposes. The methodology enables rapid estimation of inelastic moments. The methodology can easily be extended for large composite building frames where a very significant saving in computational effort would result. The feasibility of the methodology for building frames has been demonstrated by considering a single story frame.