| description abstract | In this study, a series of 1-g large shaking table tests were conducted on a soil layer with embedded twin tunnels to evaluate the acceleration distribution pattern at the ground surface. Furthermore, numerical analyses were executed to simulate the shaking table tests. The agreement between the results of experimental records and the predicted numerical results confirmed the validation of the numerical simulations. Additionally, further numerical analysis indicated that several parameters, including dimensionless period of input motion (λ/D), distance between two tunnels (s/a), embedment depth of twin tunnels (h/a), and soil shear-wave velocity (Vs), altered the peak ground acceleration pattern at the ground surface. As an important indication for λ/D = 2 and 4, the surface accelerations were attenuated between two tunnels, in contrast to λ/D = 6 and 8. Also, numerical analyses demonstrated that for small distances between two tunnels (s/a = 2.6 and 3), the twin tunnels (together) affected the amplification pattern. However, with increasing the distance between two tunnels (s/a = 4, 5, and 6), each tunnel acted separately, similar to a single tunnel. | |
