STRUCTURE OF MATERIALS
150 ESRF
Figure 132 shows a fault formed in lower crustal rocks from the Bergen Arcs in Western Norway during the Caledonian orogeny. Observations suggest that the fault caused a ca. 2 m displacement during a single large earthquake (Figure 132a). A ca. 10 cm-thick damage zone is asymmetrically developed on the northern side of the fault. This zone is interpreted to have formed prior to the actual slip on the fault, during propagation of the dynamic earthquake rupture. Within the damage zone, fragmentation is not accompanied by shear strain and the original mineral grains keep their shape. During the subsequent seismic slip, frictional heating caused melting of the wall rocks along the fault. The solidified melt is observed as a thin layer of a dark fine-grained pseudotachylyte. This layer is zoned, with an outer and inner part and a layer of sheared and fragmented wall rocks (cataclasite) separating the pseudotachylyte from the pristine wall rocks (Figure 132b). The 3D structure of this sequence of layers, was imaged by µ-CT on beamline ID19, with a voxel size of 6.5 µm and an energy of 92 keV (Figure 132c).
Fragmentation of wall rock garnets in the cataclastic layer was compared to the mechanical damage of garnets observed in
the damage zone shown in Figure 132a. Figure 133 shows EBSD orientation maps of a trail of garnet fragments from the cataclastic zone (Figure 133a) and a fragmented garnet crystal from the damage zone (Figure 133b). Both originated from a single garnet crystal in the host rock. While fragmentation within the slip zone is characterised by intense shear strain, grain rotation, and a power law fragment size area distribution with an exponent near -1.8 for most of the grains, the garnet in the damage zone of the wall rock is fragmented with negligible shear strain and grain rotation and a power law size distribution with an exponent near -2.0. This emphasises an important difference in fragmentation mechanisms within and outside the earthquake fault zone. Within the fault zone, fragmentation takes place by grinding and tear during shear. In the damage zone, however, fragmentation occurs without shear. This is often referred to as pulverisation and is believed to occur as a response to the extremely high stresses prevailing near the tip of the dynamic rupture. Because the stress field near the rupture tip is highly asymmetric and rocks are much weaker in tension than compression, this kind of damage is often only observed on the side where tension is dominant. Fragmentation due
Fig. 132: a) Field image of earthquake-induced lower crustal fault from the Bergen Arcs, Western Norway. Dark arrows show displacement directions. A damage zone of fragmented host rocks is present on the northern side of the fault and is locally intruded by pseudotachylyte (Pst) injection veins. b) Microphotograph showing a
zoned pseudotachylyte (inner and outer pst) bounded by cataclasites. The wall rock is comprised of garnet (Grt), plagioclase (Plg), and diopside (Di). c) X-ray computed microtomography image of the fault shown in (b).