Water weakening during semibrittle flow and faulting of experimentally deformed quartz sandstone

Abstract:

Triaxial compression experiments were conducted on Fontainebleau sandstone at temperatures to 900°C and effective pressures to 175 MPa with varying water contents. Both yield and peak strengths associated with semibrittle faulting decrease linearly with an increase in intragranular water concentration (C_OH); C_OH is determined from infrared spectroscopy. Microstructural observations and the influence of strain rate, temperature, and C_OH on peak strength suggest that transient semibrittle flow is accommodated through cataclasis assisted by stress-corrosion microcracking. The roles of the experimental variables on the constitutive behavior are similar to those reported for subcritical cracking of quartz single crystals. At high C_OH, microstructural observations indicate an increase in the relative contribution of mm-scale distributed shear fractures (bands) to axial strain, reflecting a reduction in grain-scale fracture toughness. This is consistent with the inference that highly dissipative shear fractures lead to the observed reduction in strength at high C_OH. Stress vs. strain rate data for transient semibrittle flow show temperature-dependent rate behavior, and are well fit by an exponential law with an activation enthalpy of 185 to 250 kJ/mol and Peierls stress of 2.5 to 7.5 GPa. Using these constraints, we infer that stress-corrosion cracking is rate-limited by the dislocation activity at crack tips. Correlation of microscale C_OH maps to microstructures suggests that intragranular water in the undeformed sandstone is associated mainly with clusters of fluid inclusions, resulting in a highly nonuniform distribution of C_OH both within and between grains. Axially deformed samples show a reduction in the median and variability of C_OH over a range of length scales. We observe that a local reduction in C_OH correlates with fluid inclusions that are decrepitated and crosscut by intragranular fractures. We conclude that intragranular fracture is the primary mechanism of water transport in our experiments.