Abstract
The predicting of the ultimate axial capacity of concrete-filled steel tube (CFST) columns is essential in the design process. To reach an accurate estimation of the column
capacity it needs to involve all the important parameters that affect the behavior. One
of the most recent tools that has been progressively used for this objective is machine
intelligence. The present study has been dedicated in order to construct an appropriate
artificial neural network (ANN) model that can give accurate predictions to the ultimate
capacity of octagonal concrete-filled steel tube columns. The model is constructed using
a back-propagation and optimizing Levenberg-Marquardt algorithms. The ANN model
depends on experimental data collected from previous researches. To show the reliability
of the model, the results were verified and compared with previous methods presented
in previous researches and codes. The ANN results showed good agreement with experimental ones. The ANN model has been used in exploring various parameters that may
affect the strength of octagonal CFST columns. The results have shown that a careful
portioning of geometrical ad material properties should be followed to achieve the most
.optimum design
capacity it needs to involve all the important parameters that affect the behavior. One
of the most recent tools that has been progressively used for this objective is machine
intelligence. The present study has been dedicated in order to construct an appropriate
artificial neural network (ANN) model that can give accurate predictions to the ultimate
capacity of octagonal concrete-filled steel tube columns. The model is constructed using
a back-propagation and optimizing Levenberg-Marquardt algorithms. The ANN model
depends on experimental data collected from previous researches. To show the reliability
of the model, the results were verified and compared with previous methods presented
in previous researches and codes. The ANN results showed good agreement with experimental ones. The ANN model has been used in exploring various parameters that may
affect the strength of octagonal CFST columns. The results have shown that a careful
portioning of geometrical ad material properties should be followed to achieve the most
.optimum design
Keywords
ANN
axial capacity
CFST
Composite
confinement