[S-SS10] Fault Rheology and Earthquake Physics

[SSS10-P27] Examination of development process of boundary shear in powdered quartz gouge by rotary-friction experiments

*Minoru Kudose¹, Tetsuro Hirono¹, Hatanaka Mamoru², Akito Tsutsumi², Takeshi Miyamoto², Satoshi Yukawa³ (1.Osaka Metropolitan University, 2.Kyoto University, 3.Osaka University)

Keywords:Friction experiment, Numerical analysis

Understanding the behavior of fault slip during earthquakes is extremely important for evaluating the amount of fault slip and estimating the energy released. In the process of faulting and increasing displacement due to shear of the subsurface rocks, plane structures with specific orientations such as Riedel and Y shear planes, develop. These are called composite planar fabric. Also, it is widely known that various foliation structures develop, not just the development of composite planar fabric, leading to thickening of shear zones. Furthermore, boundary shear is distinct from composite structures, with strain and displacement concentrated along either the hanging wall or the footwall. This has been reported through observations of internal fault structures in the field and observations of samples after friction experiments. (e.g., Marone and Scholz, 1989; Niemeijer et al. Niemeijer et al.). Numerical analysis of the shear deformation of granular materials has been conducted in previous studies, and the frictional behavior shows three stages of change ascending, descending, and steady state. Towards the end of the steady-state phase, a distinct discontinuous surface in particle velocity that correspond to boundary shear, was observed near the boundary between the gouge layer and the moving-side pressure plate. (Miyamoto et al.). However, the physical picture of the development of the boundary shear and its elementary processes are still largely unknown. Laboratory friction experiments and microstructure observations were conducted to elucidate the process of boundary shear development. Also, Numerical analysis will be conducted in parallel. In the laboratory rock friction experiments, a rotary friction tester installed at Kyoto University was used. In setting up the sample and experimental conditions, we attempted to eliminate scale effects by setting the value of the inertial number, a dimensionless quantity that qualitatively expresses the state of shear, to almost the same valueas (1e-4) in a previous study (Miyamoto et al., 2022). Specifically, artificial quartz sand with a grain size of 24.98 μm was used as the sample, with a slip velocity of 0.07 m/s and a vertical stress of 5 MPa. A sample of 2 g was used so that the thickness of the simulated fault gouge was approximately 1 mm at the start of the experiment. The maximum slip distance was 13 m. Thin sections of pre-compaction samples and samples at slip distances of 2, 6, and 13 m were prepared for observation of the deformed microstructure using a polarized light microscope. In parallel with the laboratory experiments, numerical analyses simulating the friction experiments were performed using the free software LIGGGHTS with a three-dimensional discrete element method. Measured values of Young's modulus and Poisson's ratio in quartz (Ohno et al., 2006) were set. A total of 1,700 particles with diameters of 1, 1.5, and 2 mm (average diameter 1.2 mm) were used, the slip velocity was set to 0.16 m/s, and the vertical stress was 25 MPa. The gouge thickness at the beginning of the analysis under vertical stress is approximately 17 mm. In the laboratory friction test, when slip distance was set to 13 m, localized grain refinement and discoloration were observed at the boundary surface, and this deformation was considered to correspond to a boundary shear. On the other hand, when slip distance was 2 m and the slipping was stopped just before the last sharp weakening, grain size reduction near the boundary was not observed, indicating a difference in the structure. The friction behavior of the slab decreased sharply when slip distance reached about 5 m, and remained almost constant at that low value. As a future plan, we confirmed something corresponding to boundary shear through thin section observation with a slip distance of 13m. To improve the accuracy of comparisons with numerical analysis, we will review the experimental conditions and conduct experiments under more realistic conditions. In the presentation, we will report the preliminary results of the experiment and the results of the numerical analysis, and also the results of our investigation of the relationship between the development of the boundary shear, frictional behavior, slip distance, and time transition of the particle state.