17:15 〜 18:45
[SCG40-P30] Abrupt host rock fragmentation induced by fluid phase transition during fault slip and its impact on earthquake energy partitioning
キーワード：Hydrothermal high velocity friction、Rock fragmentation、Fracture energy、Radiation efficiency
Large earthquakes typically occur at depths where temperatures are between 150-300℃; this region is called the seismogenic zone. Knowledge of the frictional behavior during coseismic slip under these conditions is essential for understanding the mechanics of such earthquakes, however, there are few experimental apparatuses globally which can impose high slip velocities under hydrothermal conditions (i.e., with high-temperature and pressurized fluid). We have developed a new apparatus, the hydrothermal high-velocity rotary shear apparatus (HDR), in Kochi Institute for Core Sample Research, JAMSTEC, which is capable of simulating coseismic fault slip under the conditions of large earthquakes in nature.
As frictional behavior under such conditions has been unexplored previously, we began by performing experiments on gabbro samples without gouge. We ran experiments under pore fluid pressures of 5-20 MPa, effective normal stresses of 5-10 MPa, temperatures of room temperature to 300 ℃, and velocities of 10 mm/s and 1.8 m/s. During the tests at high velocity conditions, we observed similar dynamic-weakening behavior at all temperatures. During the low velocity tests, we did not observe dynamic weakening but steady-state friction coefficient of ~0.6. However, in some experiments at low velocity conditions, abrupt shortening occurred after some shear displacement. We modelled the thermal evolution of the sample assembly and fluid in the pressure vessel, and found that the fluid (water in these experiments) underwent a phase transition from liquid to gas just before the abrupt shortening of the sample. The post-experiment sample was extensively fragmented, suggesting that the gas phase permeated into the rock causing it to fragment.
Interestingly, the fragmentation induced by fluid phase transition did not affect the shear strength. This suggests that the expansion of pore fluid due to the transition from fluid to gas has not effectively induced thermal pressurization during rock-rock friction, possibly due to high permeability of fault surface. Production of fine grains due to fragmentation leads to an increase in surface energy. In general, frictional work in laboratory experiments is partitioned into frictional heat and comminution. Our results indicate that phase transition-driven fragmentation consumes some portions of frictional heat. This may lead to a situation where a part of energy that should has been consumed as frictional heat may be involved in fracture energy instead. This scenario cannot be recognized from the mechanical data because fluid phase transition does not affect shear strength. The increase in fracture energy without varying slip-weakening trend can only be recognized by a decrease in radiation efficiency which results in a low rupture velocity. This implies that slow and tsunami earthquakes can be also explained by modulation of energy partitioning induced by fluid phase transition without changing frictional behavior and pore fluid pressure.
As frictional behavior under such conditions has been unexplored previously, we began by performing experiments on gabbro samples without gouge. We ran experiments under pore fluid pressures of 5-20 MPa, effective normal stresses of 5-10 MPa, temperatures of room temperature to 300 ℃, and velocities of 10 mm/s and 1.8 m/s. During the tests at high velocity conditions, we observed similar dynamic-weakening behavior at all temperatures. During the low velocity tests, we did not observe dynamic weakening but steady-state friction coefficient of ~0.6. However, in some experiments at low velocity conditions, abrupt shortening occurred after some shear displacement. We modelled the thermal evolution of the sample assembly and fluid in the pressure vessel, and found that the fluid (water in these experiments) underwent a phase transition from liquid to gas just before the abrupt shortening of the sample. The post-experiment sample was extensively fragmented, suggesting that the gas phase permeated into the rock causing it to fragment.
Interestingly, the fragmentation induced by fluid phase transition did not affect the shear strength. This suggests that the expansion of pore fluid due to the transition from fluid to gas has not effectively induced thermal pressurization during rock-rock friction, possibly due to high permeability of fault surface. Production of fine grains due to fragmentation leads to an increase in surface energy. In general, frictional work in laboratory experiments is partitioned into frictional heat and comminution. Our results indicate that phase transition-driven fragmentation consumes some portions of frictional heat. This may lead to a situation where a part of energy that should has been consumed as frictional heat may be involved in fracture energy instead. This scenario cannot be recognized from the mechanical data because fluid phase transition does not affect shear strength. The increase in fracture energy without varying slip-weakening trend can only be recognized by a decrease in radiation efficiency which results in a low rupture velocity. This implies that slow and tsunami earthquakes can be also explained by modulation of energy partitioning induced by fluid phase transition without changing frictional behavior and pore fluid pressure.