TESTING CONSTITUTIVE EQUATIONS FOR FAULT-RELATED DEFORMATION IN THE BRITTLE-DUCTILE TRANSITION

The selection of constitutive equations to characterize deformation in the “brittle-ductile transition” -- where both brittle and plastic mechanisms are active at the grain and outcrop scales -- introduces significant uncertainty to mechanical models of faulting. Because this mid-crustal interval is thought to influence the nucleation and propagation of large earthquake ruptures, constraining this region’s constitutive properties will advance knowledge of fault behavior and lead to improved seismic hazard analysis.

Strike-slip faults in the Bear Creek field area (central Sierra Nevada, CA) are characterized by both brittle and ductile features and are interpreted to have been active at ~250 MPa and as temperatures cooled from ~500°C to ~250°C. Previous work has constrained the kinematics of a contractional fault step in the Seven Gables outcrop, making this structure an excellent candidate for testing potential constitutive equations for brittle-ductile deformation. Within the step, a leucocratic dike is stretched, thinned and rotated, providing a graphic measure of the deformation. In addition, a mylonitic foliation develops within the dike and the surrounding granodiorite between the step-bounding faults.

The geometry and kinematics of the Seven Gables step provide the basis for a finite element model of the deformation, which is used to test five constitutive equations: Von Mises elastoplasticity, Drucker-Prager elastoplasticity, power law creep, two-layer elastoviscoplasticity, and coupled elastoviscoplasticity. Drucker-Prager elastoplasticity produces only elastic deformation within the step, because elevated mean normal stress in that region results in an increased plastic yield stress. Furthermore, power-law creep fails to localize strain within the step due to symmetry of the intermediate and minimum principal deviatoric stress distributions across fault planes. Of the models tested, coupled elastoviscoplasticity provides the most accurate representation of the step deformation. This model’s success is consistent with microstructural observations of brittle features (e.g., microfractures), intracrystalline plasticity (e.g., crystallographic preferred orientations), and viscous behavior (e.g., dislocation/diffusion creep) within the step.