Several well-known mechanical models have shown that unusual boundary or loading conditions can alter principal-stress orientations into configurations consistent with low-angle normal faulting. Such models, however, have not demonstrated whether magnitudes of reoriented stresses are sufficient to initiate and promote slip on low-angle surfaces. We present the results of simple Coulomb failure analyses to determine whether, and where, such models predict frictional slip, assuming geologically plausible boundary stresses, pore pressures, and rock strengths. Models that invoke a sizable shear traction at the base of the upper crust or spatially varying loads on the upper crust reorient principal stresses and failure planes but do not produce frictional failure on crustal-scale detachments either in the absence of pore fluids or at hydrostatic pore fluid pressures. Models that reorient stresses by midcrustal dike intrusion produce slip on low-angle surfaces at relatively deep crustal levels but only in the area of the dike tip; the low-angle failure surfaces curve into a high-angle orientation a short distance from the dike. All of these models also imply unsustainably high absolute tensile stresses in the upper 5 km of the crust and suggest that, in any system in which stresses are allowed to evolve over time, failure and stress release will occur on high-angle faults before low-angle ones have developed. These assertions are true even when near-lithostatic pore pressures are assumed, unless there is an inhomogeneous, extraordinarily fortuitous distribution of pore pressures and rock strengths at the time of initiation of a new detachment fault. One model we tested, for example, required pore pressures exceeding 0.96 times lithostatic in the area of the hypothesized low-angle normal fault, with lower pore pressures both above and below the detachment to prevent slip and stress release on high-angle normal faults in the upper part of the modeled region and on low-angle thrust faults in the lower part.
Xx120Times Cited:21Cited References Count:41