4:00 PM - 4:15 PM
[SVC32-08] 2D analogue modeling experiment to elucidate intrusion propagation process of high viscosity magma
Keywords:Analogue experiments, dyke intrusion, high viscosity magma
Magma propagates vertically or obliquely, as seen in the 2000 eruption of Mt. Usu and the 2014 eruption of Bardarbunga (Iceland). The direction of magma propagation may be influenced by the stress field, the magma and host rock properties, and the pressure of magma. The influence of these factors has been studied by various means, such as numerical calculations and laboratory experiments. In laboratory experiments, many experiments have been conducted in which water or air is injected into gelatin to simulate low-viscosity magma propagation. On the other hand, not many experiments have simulated high-viscosity magma propagation. In addition, experimental studies on the magma propagation process of high viscosity magma and the surface arrival position have not been carried out yet. We conducted two-dimensional analogue model experiments to identify the characteristics of the high viscosity magma propagation process under various conditions.
We injected mizuame into silica powder from the bottom of the experimental equipment to analogue the high-viscosity magma propagation. Three types of experiments were conducted. 1) homogeneous host rock and flat surface (homogeneous experiment), 2) homogeneous host rock and a mountain body on the surface (edifice experiment), 3) host rock with two layers and flat surface (layer experiment). We used two-dimensional experimental equipment with 200 mm high, 460 mm long, 50 mm thick. The silica powder's density, cohesion, and internal friction angle were 1500 kg/m3, 600 Pa, and 34 º, respectively. The density and viscosity of mizuame were 1380 kg/m3 and 11.6 Pa s, respectively. The scaling of dimensionless numbers shows that this experiment was scaled the propagation of the high viscosity magma (105-107 Pa s) from 1000 m depth.
The propagation of the mizuame went through a two-stage growing process in most of the experiments. The mizuame propagated almost vertically in the first stage, and a reverse fault was formed at the tip. The syrup propagated along the reverse fault in the second stage and reached the surface. The angle of the line connecting the injection part of the mizuame and the surface arrival position of the mizuame was used to determine the distribution of the surface arrival position.
In the homogeneous experiment, mizuame reached the point of 20-39º (Fig. 1a). This may be due to the formation of dipping reverse faults in the first stage and the propagation of the mizuame along with them. In the edifice experiment, mizuame reached the point of 0-9º (Fig. 1b). This may be due to the propagation of the mizuame along the axis of principal compressive stress toward the summit.
In almost all experiments, we confirmed cracks of silica powder's surface during the mizuame intrusion. These cracks were formed much earlier in the experiments, revealing that these formed earlier than the dike inclining. The ratio of the horizontal distance between the first and second cracks and the horizontal distance between the first crack and the surface arrival position of mizuame suggested that the mizuame tended to reach the position 1-4 times of the horizontal distance between the first and second cracks (Fig. 1c). This result suggested that the eruption position can be estimated from the crack formation process in nature eruption. However, to apply it to nature eruption, it is necessary to study in a more complicated system such as a three-dimensional experiment.
We injected mizuame into silica powder from the bottom of the experimental equipment to analogue the high-viscosity magma propagation. Three types of experiments were conducted. 1) homogeneous host rock and flat surface (homogeneous experiment), 2) homogeneous host rock and a mountain body on the surface (edifice experiment), 3) host rock with two layers and flat surface (layer experiment). We used two-dimensional experimental equipment with 200 mm high, 460 mm long, 50 mm thick. The silica powder's density, cohesion, and internal friction angle were 1500 kg/m3, 600 Pa, and 34 º, respectively. The density and viscosity of mizuame were 1380 kg/m3 and 11.6 Pa s, respectively. The scaling of dimensionless numbers shows that this experiment was scaled the propagation of the high viscosity magma (105-107 Pa s) from 1000 m depth.
The propagation of the mizuame went through a two-stage growing process in most of the experiments. The mizuame propagated almost vertically in the first stage, and a reverse fault was formed at the tip. The syrup propagated along the reverse fault in the second stage and reached the surface. The angle of the line connecting the injection part of the mizuame and the surface arrival position of the mizuame was used to determine the distribution of the surface arrival position.
In the homogeneous experiment, mizuame reached the point of 20-39º (Fig. 1a). This may be due to the formation of dipping reverse faults in the first stage and the propagation of the mizuame along with them. In the edifice experiment, mizuame reached the point of 0-9º (Fig. 1b). This may be due to the propagation of the mizuame along the axis of principal compressive stress toward the summit.
In almost all experiments, we confirmed cracks of silica powder's surface during the mizuame intrusion. These cracks were formed much earlier in the experiments, revealing that these formed earlier than the dike inclining. The ratio of the horizontal distance between the first and second cracks and the horizontal distance between the first crack and the surface arrival position of mizuame suggested that the mizuame tended to reach the position 1-4 times of the horizontal distance between the first and second cracks (Fig. 1c). This result suggested that the eruption position can be estimated from the crack formation process in nature eruption. However, to apply it to nature eruption, it is necessary to study in a more complicated system such as a three-dimensional experiment.