PRESENTING GROW, A POWERFUL NEW ALGORITHM FOR AUTOMATED FAULT GROWTH BY WORK MINIMIZATION

Laboratory experiments on brittle materials under compression demonstrate that faults grow and coalesce through a variety of failure modes and geometries. The presence of these multiple mechanisms presents a challenge to using fault growth models to constrain the evolution of fault systems. We present GROW (GRowth by Optimization of Work), a new algorithm that automates fault growth while adhering to the principals of linear elastic fracture mechanics. GROW assumes that faults evolve to maximize the mechanical efficiency of the entire system by minimizing the work (or energy) expended during deformation. This global approach accounts for the multiple mechanisms contributing to fault propagation and provides a powerful tool to study fault linkage and fault system evolution. Specifically, fault growth occurs along a pupative growth element at a fault tip that is oriented to maximize the reduction in energy of the entire system (i.e. to maximize the change in external work, ΔWext). This fault will propagate when ΔWext exceeds the energy needed to grow the fault (i.e. the work required for propagation, ΔWprop). For a 60° dipping fault, GROW predicts growth that is consistent with laboratory experiments and numerical models of fault growth that use traditional propagation criteria, producing an opening-mode wing crack at 115° clockwise from the fault under uniaxial compression and causing co-linear fault growth by shearing under biaxial compression. We analyze the propensity of faults in different orientations and spacing to grow and link under specific loading conditions, showing that the initial fault spacing controls both the migration of slip along individual faults and strain localization along a series of faults. This demonstrates the dependence of fault system development on the initial arrangement of faults and their linkage. We also test the sensitivity of fault linkage and fault system evolution to rock properties and to both static and dynamic friction.