1 INTRODUCTION Traditionally, the stability of rock slopes is assessed using limit equilibrium methods. Often these methods are limited to the study of planar or wedge instabilities and other complex failure kinematics, including toppling, are generally not addressed. It follows that the failure mechanisms are often overlooked in the assessment of the hazard of instability and, subsequently, the risk to down- slope structures. Furthermore, quantitative prediction of rock fall and runout trajectories is of extreme importance in hazard assessment of natural and engineered slopes and in the design of protective measures. To the authors’ knowledge, no commercial code is able to account for multiple modes of instability and the subsequent fragmentation, runout, and trajectories of failed material without assuming the mode or volume of failure a priori. In the current study, the ability of FEMDEM to model the onset of failure and the subsequent movement of the unstable volume is discussed. 2 THE 2D COMBINED FINITE-DISCRETE ELEMENT METHOD (FEMDEM) FEMDEM is a numerical tool pioneered by Munjiza et al. (1995) for the dynamic simulation of multiple deformable and fracturable bodies. Within the framework of FEMDEM, discrete element method (DEM) principles are used to model interaction between different solids, whose deformation is analyzed by finite element analysis (FEM). A unique feature of such a numerical tool is its capacity to model the transition from continuous to discontinuous behaviour by explicitly considering fracture and fragmentation processes. Since an explicit time-marching scheme is used to integrate Newton’s equations of motion, fully dynamic simulations can be performed. A number of different dissipative mechanisms (e.g., friction, contact damping and fracturing) are implemented into the code while no artificial numerical damping or other numerical parameters, often needed by other DEM codes, are used. Three of the key-algorithms of the method (i.e., the fracture model, the contact interaction algorithm and the dissipative impact model) are discussed in the following. For further details the reader should refer to Mahabadi et al (2011). 2.1 Fracture model A combined single and smeared strain-based crack model (Munjiza et al. 1999), also known as the discrete crack model, is implemented in the FEMDEM code used for this study. In such a model, a typical stress-strain curve for rock (in direct tension) is divided into two sections: (1) strain hardening prior to reaching the peak deformation, which is implemented in FEMDEM through a constitutive law; and (2) strain-softening where the stress decreases with an increase in strain. Material separation processes and crack process zone are modelled by means of cohesive elements as a gradual strength reduction governed by a softening law. Material strength mobilization is represented by shear and normal bonding stresses, τ and σ, which are generated by the separation of the crack edges. As depicted in Figure 1, these stresses are assumed to be a function of fracture opening, o, and sliding distance, s. Fracture behaviour is ultimately controlled by the following input parameters: tensile strength f t ; elastic modulus E; shear strength f s , which is related to material cohesion c and internal friction angle φ; and fracture energy release rate G f . Although re- meshing is not performed ahead of the crack tip, by using sufficiently fine meshes the direction of fracture propagation can be correctly captured. Slope stability analysis using a hybrid Finite-Discrete Element method code (FEMDEM) G. Grasselli, A. Lisjak, O.K. Mahabadi, B.S.A. Tatone Geomechanics Research Group, Lassonde Institute, Department of Civil Engineering, University of Toronto, ON, Canada ABSTRACT: The present paper summarizes the work done on developing and validating the use of a hybrid Finite – Discrete Element (FEMDEM) code for modelling slope instability processes. The main advantages of the FEMDEM method are the ability (i) to analyze the slope failure without assuming any mode or mechanism a priori, and (ii) to explicitly model the post-peak failure and the dynamic processes associated with the material runout. Several examples are presented to demonstrate the validity of the proposed numerical approach SUBJECT: Modelling and numerical methods KEYWORDS: numerical modelling, rock failure, rock slopes and foundations, stability analysis, rock joints, rock mass.