Prediction of Mechanical Behavior of Concrete Filled Steel Tube Structure Using Artificial Neural Network

Abstract

Artificial neural network (ANN) is applied to predict load-strain relationship of concrete filled steel tube (CFT) structural parts. An ANN prediction model, which is able to predict load-strain relationship of CFT structural parts with different dimensions and parameters, is made through training the ANN prediction model with the experimental test data. Furthermore, the prediction data and experimental test data are compared. The result shows that the combination of several characteristic parameters of CFT structural parts and ANN prediction model to predict load-strain relationship of CFT structural parts are reliable and feasible. The ANN prediction model has simple, convenience and time-saving merits.