2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 12
Presentation Time: 11:10 AM

GEOMECHANICAL MODELLING OF COMPLEX ROCK FAILURE MECHANISMS


STEAD, Douglas, Department of Earth Sciences, Simon Fraser Univ, 8888 University Drive, Burnaby, BC V5A 1S6, Canada, COGGAN, John, Camborne School of Mines, Univ of Exeter, Redruth, Cornwall, TR15 3SE, United Kingdom and EBERHARDT, Erik, Engineering Geology, Swiss Federal Institute of Technology (ETH Zurich), ETH Hoenggerberg, Zurich, CH-8093, Switzerland, dstead@sfu.ca

Modelling of rock failure has traditionally been undertaken using either continuum or discontinuum techniques. Some codes have utilized a hybrid approach where appropriate, but to date, few FEM/DEM codes have incorporated intact rock fracture capabilities. Continuum codes are able to predict failure according to a specific constitutive criterion but in general are unable to accommodate more than a few discrete discontinuities. Where discontinuity controlled deformation is important, for example in jointed rock masses, the engineer must resort to distinct element codes. These codes allow both deformation of the intact rock and displacement/opening of joints to be simulated. Some distinct element codes are based on particle code algorithms and allow the rock mass to be simulated as a series of blocks made up of circles (2D) or spheres (3D). These particles may be bonded, allowing fracture to be simulated by bond breakage between particles.

In this paper, an alternative approach is presented which uses a combined finite – discrete element code, ELFEN, which incorporates fracture propagation capabilities. The rock mass may be simulated as a jointed block medium. Using varied constitutive criteria, the development of a crack within the finite element mesh can be simulated. An adaptive meshing algorithm allows the propagation of the crack to be followed until eventually new discrete particles are formed. The author will present examples of the use of the code in modelling complex rock failures in both underground and surface rock engineering environments. The power of the models presented lies in their ability to model rock failure as a dynamic process and in particular the ability to incorporate the evolving kinematics accompanying rock failure.