Abstract. Triaxial compression experiments were conducted to investigate the inelastic and failure behavior of six sandstones with porosities ranging from 15% to 35%. A broad range of effective pressures was used so that the transition in failure mode from brittle faulting to cataclastic flow could be observed. In the brittle faulting regime, shear-induced dilation initiates in the prepeak stage at a stress level C' which increases with effective mean stress. Under elevated effective pressures, a sample fails by cataclastic flow. Strain hardening and shear-enhanced compaction initiates at a stress level C* which decreases with increasing effective mean stress.The critical stresses C' and C* were marked by surges in acoustic emission. In the stress space, C* maps out an approximately elliptical yield envelope, in accordance with the critical state and cap models. Using plasticity theory, the flow rule associated with this yield envelope was used to predict porosity changes which are comparable to experimental data. In the brittle faulting regime the associated flow rule predicts dilatancy to increase with decreasing effective pressure in qualitative agreement with the experimental observations. The data were also compared with prediction of a nonassociative model on the onset of shear localization. Experimental data suggest that a quantitative measure of brittleness is provided by the grain crushing pressure (which decreases with increasing porosity and grain size). Geologic data on tectonic faulting in siliciclasfic formations (of different porosity and grain size) are consistent with the laboratory observations.

No abstract

Abstract. Triaxial compression experiments were conducted to investigate influences of stress and failure mode on axial permeability of five sandstones with porosities ranging from 15% to 35%. In the cataclastic flow regime, permeability and porosity changes closely track one another. A drastic decrease in permeability was triggered by the onset of shear-enhanced compaction caused by grain crushing and pore collapse. The compactive yield stress C* maps out a boundary in stress space separating two different types of permeability evolution. Before C* is attained, permeability and porosity both decrease with increasing effective mean stress, but they are independent of deviatoric stresses. However, with loading beyond C*, both permeability and porosity changes are strongly dependent on the deviatoric and effective mean stresses. In the brittle faulting regime, permeability and porosity changes are more complex. Before the onset of shear-induced dilation C', both permeability and porosity decrease with increasing effective mean stress. Beyond C', permeability may actually decrease in a dilating rock prior to brittle failure. After the peak stress has been attained, the development of a relatively impermeable shear band causes an accelerated decrease of permeability. Permeability evolution in porous sandstones is compared with that in low-porosity crystalline rocks. A conceptual model for the coupling of deformation and fluid transport is proposed in the form of a deformation-permeability map.

No abstract