FLAC3D is an important numerical analysis software to study dynamic response of rock slope. In application of the program, the common issues such as absorbing boundary and the method to apply wave have been well studied. However, effects of model mesh size, damping and width of slope are still unknown. In order to make the result more convictive, this article has discussed the above issues systematically. It found that when the maximum mesh size is smaller than 1/10~1/8 of the smallest wavelength, the mesh size is smaller, the computational accuracy is higher. Meanwhile, values of damping and width of slope have no effect on the result. Additional, results of FLAC3D modeling on wave transmission across rock joints agree well with analytical solutions, so it has verified the capability and accuracy of FLAC3D on dynamic response of rock slope.
China is a mountainous country accompanied with many earthquakes. Thousands of geological disasters are triggered by the Wenchuan Earthquake in Sichuan Province, and the majority of the disasters are landslides induced (Xu et al., 2009). Meanwhile, the paper (Li et al., 2003) points out five significant issues about dynamic loading in the area of geotechnical engineering, which includes the study of rock slope stability under earthquakes, and the dynamic response of rock slopes is one of its important study issues.
The classification of rock slope stability under earthquakes has been studied in different views (Babanouri et al., 2013; Hong & Xu, 2005; Liu et al., 2001; Liu et al.,2005; Liu et al.,2007; Qi, 2002), from which we can conclude that the classification mainly includes pseudo-static, sliding block, numerical, model test and probability analysis methods etc. Meanwhile, the analysis method of dynamic response of rock slopes can be classified as analytical, numerical and model test methods. However, analytical solutions encounter serious limitations when the problem geometry is complicated and a great number of influencing parameters are to be incorporated (Gong, 2000; Qi, 2006). And results obtained by model test and in-situ test methods are limited to be generalized. Due to complicated topography and slope structure, results from in-situ test are difficult to obtain and the conclusion from model tests cannot reflect the natural slope. While numerical simulation is an economical and convenient alternative method. Until now, a variety of numerical simulation methods have been applied in dynamic analysis successfully, including finite element, finite difference, discrete element, fast Lagrange element, discontinuous deformation analysis, manifold element, boundary element, unbounded element and semi-analytical element methods (Qi et al., 2004). Nowadays, the most popular methods are finite element, discrete element and fast Lagrange element methods at home and abroad (Liu et al., 2007). And fast Lagrange element method follows the continuum hypothesis, which adopts difference format. With continual change of the configuration, the coordinates update according to the solution of time step integral. Based on the dynamic equation and dynamic method, it is able to simulate the dynamic problem well. Recently, FLAC3D (Fast Lagrangian Analysis of Continua in 3 Dimensions) which is based on this principle has been very popular in the area of geotechnical engineering.
