MODELING THE CAPACITY OF PIN-ENDED SLENDER REINFORCED-CONCRETE COLUMNS USING NEURAL NETWORKS

This study demonstrates the feasibility of using multilayer feedforwar d neural networks to model the complicated nonlinear relationship betw een the various input parameters associated with reinforced concrete c olumns and the actual ultimate capacity of the column. The neural netw ork models were constructed directly from a fairly comprehensive set o f experimental results and were found to be tolerant of certain levels of errors in the original testing results. Comparison with the origin al testing data and theoretical model showed that the ultimate capacit y of reinforced concrete columns predicted by the neural network model s is reasonably accurate. Parametric analysis indicates that the neura l network model has reasonably captured the behavior of reinforced con crete columns. Numerical studies are conducted to investigate modeling issues such as different data scaling schemes and dimensionless repre sentation schemes. Nonlinear transformation of the output values resul ted in an overall improvement in the generalization capabilities of th e neural network model, Preliminary studies using a limited data set o f 54 test results on high strength concrete columns also showed promis ing results. The neural network model can be useful in checking routin e designs because it provides instantaneous results once it is properl y trained and tested.