Seismic response of the slope under different boundary conditions through shaking table test

The physical slope models are commonly constructed in rigid or laminated shear model containers during shaking table tests. To reduce the effect of the boundary on the seismic response of the slope, the boundary treatment material is often filled between the physical model and model container or the slope model is lengthened. In this study, damping fluid was applied for boundary treatment and five homogeneous slopes with the same slope geometry were considered, while different lengths (M_1 and M_2 are larger and smaller models, respectively) and boundary materials. The AAF of M_1 and M_2 increased with elevation, while M_3 fluctuated and M_4 had a stronger amplification. The acceleration amplification factors at the leading edge of models M_1 and M_2 are lower than those of the other models, with slight differences in peak ground acceleration. The coherence exceeded 90% within the dominant frequency range of the input seismic waves of M_1 and M_2. Under low-amplitude (≤0.5 g) and low-frequency (≤50 Hz) excitation, small model incorporating the newly developed boundary material can effectively reproduce the seismic response of slopes while enlarged model dimensions are required under strong and higher-frequency excitation. This study provides an effective boundary treatment method for shaking table tests of slopes.