Kinematically-equivalent but geomechanically-different simulations of fault evolution: the role of loading configurations

Geomechanical simulations are used to demonstrate the importance of the way that models are loaded. In this paper the development of permanent damage during faulting using frictional-slip models of a reverse fault is investigated. Although the use of different loads and constraints can produce the same faulted geometry (for the same rock type, and at the same burial depth), the models develop very different stress and strain states. Permanent strain magnitudes and distributions between models are quite dissimilar, including the distributions of permanent dilation and compaction. This work demonstrates that boundary loads and boundary constraints are significant factors in determining what stress and deformation states evolve in the simulation model. The examples also illustrate that final (deformed) geometry alone is a very poor basis from which to predict either stress state or open fracture distribution. Bulk finite strain does not allow a prediction of local principal stress directions, magnitudes, or signs, at least in the vicinity of fault damage zones.