Fabric memory during repeated reversal shear in clay-rich materials: An experimental study using SEM, SP-µCT, and AMS

Kiichiro Kawamura 

Fine-grained shear zones occur in both tectonic faults and gravitational landslides, making their origin difficult to determine from individual microstructural features alone. A potentially useful distinction is provided by deformation history: tectonic faults commonly experience repeated earthquake cycles in which the sense and magnitude of deformation may change, whereas landslide slip zones are primarily governed by gravitational downslope loading. How reversal of shear is recorded in clay-rich microfabric, however, remains poorly understood. Here, I investigate the development of fabric memory during repeated reversal shear using direct-shear experiments on clay-rich materials from the Jin’nosuke-dani landslide, central Japan. Undisturbed and slurry samples were tested under normal stresses of 0.4, 0.8, and 1.6 MPa. Following 1000 min of pre-consolidation, specimens underwent ten predetermined shear cycles, with the shear direction reversed after each cycle. Shear stress and vertical displacement were recorded automatically, and the experimentally deformed materials were examined by scanning electron microscopy (SEM), synchrotron radiation micro-X-ray computed tomography (SP-µCT), and anisotropy of magnetic susceptibility (AMS). SEM and SP-µCT reveal preferred clay-flake orientations, fragmented and locally rotated shale clasts, asymmetric structures, and three-dimensional reorganization of the clay-rich material. AMS provides a complementary bulk-scale constraint, with Kmax approximately horizontal and oriented perpendicular to the imposed shear direction. The transverse Kmax orientation, together with the observed particle and clast rotation, is consistent with rotation and rolling of gouge-like material during reversal shear. We propose that reversal does not simply erase the fabric generated during the preceding shear cycle. Instead, elements of the earlier fabric are retained while being rotated, disrupted, and overprinted by deformation in the opposite direction. Repeated reversal therefore produces a composite fabric that records cumulative deformation history, which I refer to as experimental fabric memory. This rotational fabric memory may be associated with deformation histories involving repeated reversal and could provide a process-based line of evidence for distinguishing tectonic shear zones from gravitational landslide slip zones when combined with independent geological, geomorphological, and kinematic evidence. The present study does not establish that the observed fabric is uniquely tectonic, because the experiments were performed on landslide-derived material and the fabric was measured only after the ten-cycle test. Rather, it demonstrates a testable mechanism by which reversal deformation can leave a distinctive microfabric signature.