中文摘要：

潜在滑动面的确定和模拟是研究边坡稳定性的关键问题，滑动面的确定受多种因素的影响，较为均质的土质边坡可通过理论方法直接确定，非均质的岩质边坡均需通过多种手段和途径综合确定。目前常用的模拟滑动面模型有：无滑动面模型、薄层实体单元模型、界面单元模型。通过深入剖析各种方法的优缺点，指出无滑动面方法不能反应滑体滑动机制和失稳过程；薄层实体单元相当于用连续变形的岩体单元来模拟发生不连续变形的结构面；Goodman节理单元的法向刚度K与切向刚度怒很难确定，且易发生单元嵌入现象；Katona内界面单元或摩擦一接触型界面单元则不需要确定节理、断层的‰，整，直接采用它们的黏聚力和内摩擦角，通过接触面两侧的模量变化来反映％和眨的影响，而且可以很好地模拟接触面的滑动、张开与闭合等状态。最后借助经典工程实例对推荐的摩擦一接触型界面单元进行验证。结果表明：采用摩擦一接触型界面单元模拟滑动面的分析结果与工程实际情况吻合较好，符合边坡变形规律和工程经验，特别是结构面的模拟较为真实地反应了实际边坡工程的问题。

英文摘要：

Both the determination and simulation of the sliding surface of slopes are the key problems in the stability analysis and design of slopes. The determination of the sliding surface is affected by many factors. The sliding surface of the heterogeneous rock slope needs to be investigated by the combination of geological and theoretical methods. At present, the commonly used simulation models of the sliding surface are the non-slip surface model, the thin layer solid element model and the interface element model. It was found out that the sliding mechanism for the instability processes of a landslide could not be simulated by the non-slip surface method. In the method of thin-layer solid element, the continuous deformation of rock mass elements is used to simulate the discontinuous deformation of structure surface. As to the joint element of Goodman, it is difficult to obtain the normal stiffness Kn and the tangential stiffness Ks, and the element embedding and overlapping are produced in the finite element analysis. Nevertheless, in the Katona internal interface element or the friction-contacting interface element, the strength parameters of joints or faults can be directly adopted without the need to determine the Kn, Ks parameters. Further more, the state of sliding, opening and closing of the contact surfaces can be well simulated with the friction-contacting interface element. The element type proposed was applied to a typical engineering case in order to verify the effectiveness of the method. The friction-contacting interface element was demonstrated to be reasonable by comparing with the results from field measurements.