Author : Frans Aben
Publisher :
Page : 0 pages
File Size : 25,12 MB
Release : 2016
Category :
ISBN :
Earthquakes along large crustal scale faults are a huge hazard threatening large populations. The behavior of such faults is influenced by the fault damage zone that surrounds the fault core. Fracture damage in such fault damage zones influences each stage of the seismic cycle. The damage zone influences rupture mechanics, behaves as a fluid conduit to release pressurized fluids at depth or to give access to reactive fluids to alter the fault core, and facilitates strain during post- and interseismic periods. Also, it acts as an energy sink for earthquake energy. Here, laboratory experiments were performed to come to a better understanding of how this fracture damage is formed during coseismic transient loading, what this fracture damage can tell us about the earthquake rupture conditions along large faults, and how fracture damage is annihilated over time.First, coseismic damage generation, and specifically the formation of pulverized fault damage zone rock, is reviewed. The potential of these pulverized rocks as a coseismic marker for rupture mechanisms is discussed. Although these rocks are promising in that aspect, several open questions remain.One of these open questions is if the transient loading conditions needed for pulverization can be reduced by progressively damaging during many seismic events. The successive high strain rate loadings performed on quartz monzonites using a split Hopkinson pressure bar reveal that indeed the pulverization strain rate threshold is reduced by at least 50%.Another open question is why pulverized rocks are almost always observed in crystalline lithologies and not in more porous rock, even when crystalline and porous rocks are juxtaposed by a fault. To study this observation, high strain rate experiments were performed on porous Rothbach sandstone. The results show that pervasive pulverization below the grain scale, such as observed in crystalline rock, does not occur in the sandstone samples for the explored strain rate range (60-150 s-1). Damage is mainly occurs at a scale superior to that of the scale of the grains, with intragranular deformation occurring only in weaker regions where compaction bands are formed. The competition between inter- and intragranular damage during dynamic loading is explained with the geometric parameters of the rock in combination with two classic micromechanical models: the Hertzian contact model and the pore-emanated crack model. In conclusion, the observed microstructures can form in both quasi-static and dynamic loading regimes. Therefore caution is advised when interpreting the mechanism responsible for near-fault damage in sedimentary rock near the surface. Moreover, the results suggest that different responses of different lithologies to transient loading are responsible for sub-surface damage zone asymmetry.Finally, post-seismic annihilation of coseismic damage by calcite assisted fracture sealing has been studied in experiments, so that the coupling between strengthening and permeability of the fracture network could be studied. A sample-scale fracture network was introduced in quartz monzonite samples, followed exposure to upper crustal conditions and percolation of a fluid saturated with calcite for several months. A large recovery of up to 50% of the initial P-wave velocity drop has been observed after the sealing experiment. In contrast, the permeability remained more or less constant for the duration of the experiment. This lack of coupling between strengthening and permeability in the first stages of sealing is explained by X-ray computed micro tomography. Incipient sealing in the fracture spaces occurs downstream of flow barriers, thus in regions that do not affect the main fluid flow pathways. The decoupling of strength recovery and permeability suggests that shallow fault damage zones can remain fluid conduits for years after a seismic event, leading to significant transformations of the core and the damage zone of faults with time.