INVESTIGATING THRUST FAULT BRANCHING AND PROPAGATION USING WORK OPTIMIZATION

Numerical models of the evolution and interaction of faults at various developmental stages and physical experiments that are scaled to represent crustal deformation allow us to predict the behavior of crustal thrust faults. This research investigates thrust fault propagation and branching using the Boundary Element Method (BEM) code, Growth by Optimization of Work (GROW). The open-source GROW code applies a global work criterion to forecast fracture propagation patterns. The new GROW implementation of adding new fault branches at points where high shear stress adjacent to faults exceeds the intact shear strength enables more realistic simulation of thrust fault evolution. The two-dimensional plane strain numerical models simulate clay experiments conducted by Bonnano et al. (2017) of 30˚ and 45˚ thrust fault propagation, both with and without a superjacent horizontal discontinuity that represents a weak contact. The previous laboratory experiments showed sequences of branched and curved thrust faults with new fault branches in higher structural position accommodating slip at the expense of the older branches.

To validate the new branching implementation within GROW, we compared the numerical simulation outcomes with experimental observations, focusing on the evolution and geometry of branched faults and their interaction with pre-existing discontinuities. The numerical simulations produce fault branches and shifts in fault activity; slip along faults of lower structural position is abandoned in favor of slip along the newer and higher fault branches.

The strong agreement between the numerical results and that of the clay experiments demonstrates the feasibility of GROW’s work optimization with fault branching for simulating complex thrust fault evolution. By incorporating intricate fault branching geometries into the numerical models, we can more accurately replicate fault propagation patterns observed in the Earth's crust, where major faults often have multiple segments.