Concrete-filled steel tubes (CFSTs) have been extensively employed in pressurized structural members, such as high-rise buildings, large-span bridges or offshore structures. Accurately and reliably assessing the damage of CFST column subjected to explosion is helpful for designers to optimize the structural performance and construction cost. Currently, peak displacement, the damage indicator, is generally estimated via finite element method (FEM) and single degree of freedom (SDOF) oscillator model. However, the computational overhead of FEM is high, while the computational accuracy of SDOF model is low. In this paper, artificial neural network (ANN) is used to overcome the above drawbacks. A dataset of 270 experiments and numerical simulations is collected from existing literatures to develop the ANN model. The performance of this model is compared with the conventional methods, i.e., FEM and SDOF model. And its prediction accuracy and efficiency are better in the comparison by experimental results. The damage evaluation and parameter research are carried out to reveal insight into the factors on the CFST column under explosion, including the property of column and its cross-section properties, the axial compression condition, and the explosive. To support the explosion resistance design of a CFST column, an explicit equation and an excel calculation table are introduced based on the ANN model.
Read full abstract