The origin of S- C mylonites and a new fault-zone model

Abstract

This paper presents a synthesis of experimental data on mechanical behaviour and deformation textures of simulated halite shear zones with special regard to the internal structures of S- C mylonites and their mechanical implications. Halite is the only mineral so far for which a complete transition between brittle shearing deformation to fully ductile shearing flow in the dislocation glide regime has been studied experimentally under large shear strains that may be encountered along faults or plate boundaries. Moreover, deformation textures in the shear zones are similar to those of mylonites and cataclasites. Thus the experimental results provide an insight into the petrogenesis of fault rocks, mylonites in particular, and the establishment of a realistic fault model. Between the brittle regime and the fully ductile regime (i.e. pressure-insensitive inelastic regime), there exists a wide and distinct regime called ‘semiductile’ in which deformation textures are nearly identical to those developed in the ductile regime, yet the shear resistance is pressure-dependent, and potentially unstable fault motion can occur in the low pressure part. In cooking salt shear-zone experiments, the characteristic texture in the semiductile and ductile regimes is foliation resembling mylonitic foliation, which formed in the direction of maximum elongation. In the semiductile regime, nearly homogeneous shearing deformation, characterized by the formation of uniform and pervasive foliation, changes into heterogeneous deformation after a critical shear strain and this strain localization induces potentially unstable fault motion. The primary mode of internal slip after the strain localization is slip along Y Riedel shears, parallel to the shear-zone boundary, interconnected with R 1 and some P Riedel shears. These shear surfaces of concentrated deformation are very similar to the internal structures of so-called ‘ S- C mylonites”. Hence, the semiductile regime is the primary candidate for the formation of S- C mylonites, and at least some S- C mylonites must have formed at seismogenic depths. This interpretation is consistent with recently reported interlaced pseudotachylytes and mylonites. It is also shown that the size of recrystallized halite grains (presumably mostly postdeformational) decreases with increasing shear strains, but not with the shear stress, and this is consistent with the grain-size reduction with strain concentration which has been confirmed for many mylonites in the world. Thus, estimation of flow stress from recrystallized grains needs to be done with great care and might often be erroneous. Existing fault models are critically examined from the viewpoint of the halite experiments, and a new fault model is proposed whereby fault zones are divided into brittle, semibrittle, semiductile and ductile regimes with increasing depth. The shear resistance of faults is highest in the semiductile regime, and the seismogenic depth is likely to extend down to the upper part of this regime. The semiductile regime thus emerges as an area of primary interest with regard to the petrogenesis of many S- C mylonites, modelling of large and great shallow earthquakes along pre-existing faults and plate boundaries, and quantitative evaluation of the mechanical interactions of plates across their boundaries.