Observational studies have shown that faults are complex structures with a "core" of fine-grained cataclasites bordered by layers of gouge and breccia sandwiched between zones of damaged wall rock extending for hundreds of meters from the core (Biegel and Sammis, 2004). Key questions to answer are: (1) how does this fault zone structure form, and (2) does it affect the mechanics of the earthquake rupture? Recent theoretical studies by Rice and Sammis have shown that gouge and breccia layers can form in the dynamic stress field near the tip of an earthquake rupture, and Andrews has demonstrated that energy loss in a damage zone can significantly contribute to the fracture energy, thereby affecting the rupture velocity of an earthquake. By formulating an analytical dynamic slip-pulse model and fitting it to observational data from Heaton, Rice and Sammis were able to show that the off-fault stress field around a propagating earthquake rupture is capable of producing a gouge-breccia layer tens of meters wide and a damage zone hundreds of meters wide. While the model can predict the effect of rupture velocity on off-fault stresses, it cannot predict the extent to which off-fault damage controls the rupture velocity. In this study, The investigators will explore the feedback of active damage on rupture velocity in the laboratory. They will nucleate a series of dynamic mode II shear ruptures on the interface between two elastic plates, and use high-speed digital photography to measure rupture speed and observe the pattern of associated off-fault damage. These experiments will be carried out in collaboration with Professor Ares Rosakis in his laboratory at Caltech. His experiments in photoactive Homalite did not generate off-fault damage because there are no flaws in Homalite from which it can nucleate. The PIs will extend the Rosakis experiments by fabricating samples that have a homogeneous distribution of starter flaws to allow the nucleation of off-fault damage. The results of these experiments will be compared with predictions from the slip pulse model in which the micromechanical damage model of Ashby and Sammis (1990) will be used to calculate the extent of off-fault damage in the theoretical stress field. This comparison will serve as a reality check on the theoretical models, and allow a quantitative assessment of the feedback of active off-fault damage on the propagation of a dynamic earthquake rupture. Broader impacts include producing images for educational purposes and the support of a new PI, as well as fostering a collaboration between USC and Aeronautical Engineering Labs at Caltech.
