These observations pose several mechanical problems: Subvertical faults imply initial tensile failure to depths of at least 5 km (the minimum structural thickness of the exposed tilted fault blocks), whereas increased confining pressures should promote shear failure. Large (> 70¡) rotations on a single set of faults defies the tendency to form new faults as old faults become progressively less favorably oriented and frictional resistance to slip increases. Finally, the rapid rates of local extension imply an instability or runaway phenomena. In order to address these issues, we investigated the mechanics of extensional failure and fault block rotations from the perspective of conventional Mohr failure envelopes. We assumed s1 = vertical = rgh and derived the criteria for initial failure and subsequent frictional sliding at the base of a brittle layer (i.e. its strongest part) as a function of: initial layer thickness (h0), differential stress (sD), pore fluid pressure (Pf), tensile strength (T), cohesion (C), coeff. of friction (m), and amount of rotation (r). For nominal model values of h0 = 5 km, T0 = -10 MPa, C0 = 30 MPa, m = .58, we find: Initial tensile failure requires relatively small differential stresses (~ 40 MPa) but elevated pore fluid pressures (l = Pf/rgh = ~0.8) - not an unreasonable value - particularly within a magmatic/hydrothermal system. For the tensile fractures to then slip in shear reguires either a slight reorientation of the principal stress axis or an infinitesimal rotation of the fracture plane - either case likely dictated by the lower boundary conditions. As faults rotate Òdomino-styleÓ from 90¡ to progressively lower angles, they become progressively weaker ( i.e., require smaller sD to induce slip) at any given fluid pressure ratio. Surprisingly, this progressive weakening continues even after faults have rotated past the ÒoptimumÓ Coulomb failure angle of 60¡ to fault dips of only 30-40¡, after which they remain weak until either they reach their angle of repose (~15¡) or it becomes easier to form a new tensile fracture (see diagram). This progressive weakening stems from the fact that the increasing proportion of normal stress acting on the fault plane is outweighed by the overall reduction in mean normal stress due to thinning of the overburden and would only apply for fairly rapid strain rates (thinning of the brittle layer outpaces conductive relaxation of the isotherms). These simple mechanical considerations provide a general explanation for the striking differences in structural style between rapidly extending (± magmatically active) corridors of focused extension and areas of more broadly distributed slow extension and sheds light on why normal faults that continue to move at low angles do no typically produce large earthquakes.
|