11:00 AM - 11:15 AM
[SCG46-08] Variation of stick-slip behaviors controlled by dehydration kinetics of simulated gypsum gouges
Keywords:stick-slip, subduction zone, dehydration instability, pore fluid pressure, kinetics
In the subduction zone, the increase in pore fluid pressure causes slow slip (Katayama, 2016), slow slip accumulates stress in the asperities and accumulation of stress eventually leads to the occurrence of large earthquakes (Hasegawa et al., 2015). Because gypsum undergoes phase transition and dehydration at relatively low temperatures and pressures, it can easily reproduce the dehydration instability of hydrous minerals in the laboratory (Brantut et al., 2011). Therefore, there have been many studies on the effect of pore fluid pressure on slip events using gypsum (e.g., Leclère et al., 2016). There are also many studies on the dehydration reaction of hydrous minerals from the viewpoint of reaction kinetics (e.g., Liu et al., 2015). In order to illuminate how the dehydration kinetics controls the frictional behavior, Sasaki et al. (2016) conducted friction experiments of simulated gypsum gouges under various temperature and pressure conditions, and found the change in recurrence intervals of repeated stick-slip events with time in an experiment conducted under dehydration condition. Later by analyzing the data of Sasaki et al. (2016), Iwasaki et al. (2017) derived the time function of pore fluid pressure causing the change in the recurrence intervals assuming that gypsum underwent phase transition and dehydration during the friction experiment. However, both studies did not confirm the presence of a dehydration product (anhydrite) after experiments. In this study, we first discussed the possibility of dehydration of gypsum in the experimental conditions and then evaluated the time function of pore fluid pressure.
Sasaki et al. (2016) conducted friction experiments using gypsum gouge under the conditions of confining pressures of 10 - 200 MPa and temperatures of room temperature - 180 °C by gas apparatus. The data showed that the stress drops and recurrence periods of stick-slip events increased with increasing confining pressure at room temperature and 70 °C, while the stress drop and recurrence period decreased with time at 200 MPa and 110 °C (Iwasaki et al., 2017). This is presumably because the phase transition and dehydration of the gypsum created pore fluid pressure that reduced the apparent mechanical strength. Iwasaki et al. (2017) obtained a time function of pore fluid pressure by fitting the reaction kinetics using ‘Avrami's equation’ (Avrami, 1940) to temporal variation of pore fluid pressure.
As for the dehydration potential of gypsum, under the condition where the stress drop varied with time, the frictional coefficient decreased with time when the pore fluid pressure was not taken into account. However, taking the pore fluid pressure into account, the frictional coefficient remained almost constant, which is consistent with the previous studies reporting the similar frictional coefficients of gypsum and anhydrate. In addition, the microstructure under the presumed dehydration condition did not show much Riedel shear planes, which may be due to the decrease in shear stress caused by the occurrence of pore fluid pressure (Leclère et al., 2016). Furthermore, we calculated the porosity of the gypsum gouge on the assumption that pore fluid pressure was generated, and the value was slightly smaller than the value known from previous studies, but it was considered that dehydration reaction occurred. This suggests that the gypsum may have undergone phase transition and dehydration during the shear deformation experiment, and then we evaluated the time function of the pore fluid pressure. The Avrami equation used for the temporal variation of pore fluid pressure includes the reaction rate k and the exponent n, which is related to the formation and growth mechanism of the material. The reaction rate k depends on the conditions of temperature, pressure, etc., and is generally harmonic when compared with the conditions of previous studies (Liu et al., 2015; Ballirano and Melis, 2009; Carbone et al., 2008). The exponent n indicates the formation and growth mechanism of the material. Therefore, the obtained time function of pore fluid pressure is considered to be reasonable. If the time function of the pore fluid pressure obtained from the friction experiment using gypsum can be used to estimate the time until slow slip occurs, it will contribute to the estimation of the period of stress accumulation in the asperities and to the study of the occurrence of large earthquakes caused by slow slip.
Sasaki et al. (2016) conducted friction experiments using gypsum gouge under the conditions of confining pressures of 10 - 200 MPa and temperatures of room temperature - 180 °C by gas apparatus. The data showed that the stress drops and recurrence periods of stick-slip events increased with increasing confining pressure at room temperature and 70 °C, while the stress drop and recurrence period decreased with time at 200 MPa and 110 °C (Iwasaki et al., 2017). This is presumably because the phase transition and dehydration of the gypsum created pore fluid pressure that reduced the apparent mechanical strength. Iwasaki et al. (2017) obtained a time function of pore fluid pressure by fitting the reaction kinetics using ‘Avrami's equation’ (Avrami, 1940) to temporal variation of pore fluid pressure.
As for the dehydration potential of gypsum, under the condition where the stress drop varied with time, the frictional coefficient decreased with time when the pore fluid pressure was not taken into account. However, taking the pore fluid pressure into account, the frictional coefficient remained almost constant, which is consistent with the previous studies reporting the similar frictional coefficients of gypsum and anhydrate. In addition, the microstructure under the presumed dehydration condition did not show much Riedel shear planes, which may be due to the decrease in shear stress caused by the occurrence of pore fluid pressure (Leclère et al., 2016). Furthermore, we calculated the porosity of the gypsum gouge on the assumption that pore fluid pressure was generated, and the value was slightly smaller than the value known from previous studies, but it was considered that dehydration reaction occurred. This suggests that the gypsum may have undergone phase transition and dehydration during the shear deformation experiment, and then we evaluated the time function of the pore fluid pressure. The Avrami equation used for the temporal variation of pore fluid pressure includes the reaction rate k and the exponent n, which is related to the formation and growth mechanism of the material. The reaction rate k depends on the conditions of temperature, pressure, etc., and is generally harmonic when compared with the conditions of previous studies (Liu et al., 2015; Ballirano and Melis, 2009; Carbone et al., 2008). The exponent n indicates the formation and growth mechanism of the material. Therefore, the obtained time function of pore fluid pressure is considered to be reasonable. If the time function of the pore fluid pressure obtained from the friction experiment using gypsum can be used to estimate the time until slow slip occurs, it will contribute to the estimation of the period of stress accumulation in the asperities and to the study of the occurrence of large earthquakes caused by slow slip.