Recent experimental insights into dynamic fracture and fragmentation in fault damage zones

Inelastic yielding in fault damage zones during rupture propagation has important implications for the earthquake energy budget, static and dynamic triggering, seismic radiation, and fault zone permeability. Fault damage zones develop a variety of processes over many earthquake cycles, including process zone damage during fault growth, wear due to slip at geometric irregularities, and earthquake rupture propagation, but sorting out the exact source of damage in the field can be difficult. Because the mechanics of fracture is highly dependent on strain rate, the task of synthesizing information from rock mechanics experiments, fracture mechanics theory, and field observations of fault zone damage highly non-trivial. This challenge is typified by the enigmatic observation of pulverized fault damage zones, in which highly fragmented or "pulverized" rocks occur in asymmetric damage zones up to 100m away from the principal slip zone of seismogenic faults. Laboratory uniaxial compression experiments aimed at simulating the formation of pulverized rocks require strain rates of over 10² s^-1 and differential stresses over 10² MPa, both orders of magnitude larger than expected at large distances from faults during earthquake rupture. Further confusing the issue, pulverized rocks tend to occur preferentially in strong crystalline rocks, whereas adjacent weaker granular rocks often appear largely undeformed. Such perplexing observations beg the question whether the experiments from which we derive our understanding of failure processes are based on faulty assumptions, poorly designed, or simply difficult to scale to the natural prototype. In this talk I will summarize some recent experimental work designed to better mimic natural damage processes at fast strain rates. In particular, I will explore the mechanics of rock failure during of transient coseismic tensile stress perturbations, the role of complex loading paths on rock strength, dynamic mixed mode fracture propagation, and the influence of grain-scale structure on the style of dynamic fracture preserved in the rock record.