Seismic behavior analyses of reinforced tunnel with steel fiber reinforced concrete (SFRC) composites by consideration of optimum thickness in artificial neural network (ANN) analyses

This study is organized in two parts. In the first part, the seismic behaviour of tunnels reinforced with steel fibre-reinforced concrete composites (SFRC) fibre reinforcing is investigated and using the finite element method (FEM) in ABAQUS software, ۳ different models of the thickness (similar in terms of compressive strength and tensile strength) have been evaluated in the Tehran Metro tunnel. In this section, to apply seismic force, the PGA characteristics of the El Centro earthquake in the United States have been used, and the parameters of force and displacement concerning time have been extracted. In the second part, using the effect of three different thicknesses of ۲۰, ۴۰ and ۶۰ cm in determining the behaviour of the tunnel under the earthquake, the amount of stress and displacement over time was extracted. These results were investigated using two models of neural networks, MLP and RBF, to determine the optimal thickness to withstand stress and displacement in the tunnel. It was also found that the optimal thickness of the tunnel was obtained using MLP and RBF networks to control stress of ۴۷.۹۵ cm and ۴۸.۲۲ cm, respectively, and to control displacement of ۴۷.۲۱ cm and ۴۸.۱۵ cm, respectively. Finally, it can be acknowledged that using the optimal thicknesses, the maximum stress tolerance by MLP and RBF methods is equal to ۱.۷۵۵۶ \times ۱۰^{+ ۶}\ N/m^{۲} and ۱.۹۲۸۹ \times ۱۰^{+ ۶}\ N/m^{۲} is obtained. Also, the highest amount of displacement with both methods (due to the good accuracy of both models in determining the displacement) is equal to ۰.۰۲۵۶ cm.

کلیدواژه ها:

Seismic Analyses ، Reinforced Tunnel ، Thickness ، Neural Network