1 ABSTRACT
We have completed the first steps in attempting to validate a material constitutive model for a jointed rock mass. The continuum model (Thomas, 1981), as utilized here within a finite element code, consists of a material constitutive description based on the linear elastic behavior of the matrix material and nonlinear normal and shear behavior of fracture planes. The validation exercise consists of (1) characterizing an appropriate physical model, (2) measuring the mechanical response of that physical model, and (3) using the physical model characteristics to calculate the mechanical response of the physical model. Calculated and measured responses for small-scale physical models (thermally-fractured granite [Bauer, 1980]) are in qualitative agreement. For deviatoric loadings, the calculated tangent modulus is an order of magnitude less than that of the matrix and exhibits stiffening with increasing a load, and thereby functionally tracks the measured response.
2'INTKODUCTION
The geologic disposal of radioactive waste in the volcanic tuffs under Yucca Hourirain, Nevada, is being considered by the Nevada Nuclear Waste Storage Investigations Protect, administered by the Nevada Operations Office of the U.S. Department of Energy. Meaningful calculations that address site evaluation, repository design, and performance assessment require a constitutive description that realistically models the actual physical situation. Extensive field and drill hole data at Yucca Hourirain (Scott et al., 1983) suggest that the material model should incorporate the mechanical response of both the intact rock (matrix) and the fractures. To this end, Thomas (1981) presented a general three-dimensional model for fractured media which is composed of two parts: (1) a continuum- based technique to average the discontinuous displacements across fracture planes within a representative elementary volume (based on Hotland, 1974a; 1974b; 1977) and (2) a constitutive description based on the linear elastic behavior of the matrix material and nonlinear behavior of the fractures. Validation of such a model is a nontrivial task. The continuum assumption is realistically applicable to certain classes of field problems at Yucca Mountain. These problems range from emplacement hole to far- field simulations. On these scales we assume that we can smear out and average the contributions of the two separate mechanisms to the total deformation. Field tests are often used as part of the validation for this type of constitutive model. In designing these tests care must be taken to insure that the volumes of rock deformed are representative of the rock mass. Validation of this type of material model may also require implementation of specially designed, large-scale laboratory tests (Einstein & Hirshreid 1973). Our approach has been to use a small-scale physical model of a jointed rock mass for preliminary validation.
3 CONSTITUTIVE MODEL
The constitutive model for jointed rock masses used in this study was reported by Thomas (1981). It takes a continuum approach in the sense that every material point in the model behaves as would a representative elementary volume composed of a matrix material containing a suitably large number of joints. The total strains are decomposed into contributions from the matrix and joints, so that load sharing takes place. Normal and shearing motions of joints are related to the conjugate stresses through the stiffness matrix.
