11:00 AM - 1:00 PM
[SCG49-P05] Theoretical and experimental study of the intergranular cracking in partially molten rock
Keywords:Anelasticity, Intergranular cracking, Partially molten rock
Previous studies have suggested that grain boundary sliding is the primary mechanism of rock anelasticity, causing seismic wave dispersion and attenuation in the upper mantle. Yamauchi and Takei (2016) demonstrated that anelastic behavior is determined by the atomic-scale grain boundary structure (degree of disorder) and grain boundary diffusivity and that the presence of less than 1% melt, expected in the upper mantle, has little effect. However, the presence of melt reduces the effective confining pressure and may cause intergranular cracking under differential stress, the effect of which on anelasticity has not been investigated in previous studies. Intergranular cracking can promote grain boundary sliding, resulting in further seismic wave velocity reduction and increased attenuation. Therefore, it is very important to clarify the effect of melt on the intergranular cracking. This study aims to clarify the conditions under which intergranular cracking occurs in partially molten rocks in the upper mantle.
In this study, we theoretically considered the condition for the intergranular cracking in partially molten rock. First, we examined the applicability of the Navier-Coulomb criterion, which states that the shear stress needed for fracture is proportional to the normal compressive stress. This criterion is explained by the increase in the real contact area in proportion to the normal compressive stress applied to a rough surface. Since such a rough surface does not exist in the partially molten rocks, we concluded that the Navier-Coulomb criterion is not applicable. Therefore, the effective confining pressure theory based on the Navier-Coulomb criterion is not applicable, either. Next, we applied the Griffith theory to the crack extension at the junction of grain boundary and two solid-liquid interfaces in the partially molten system. We found that the condition for the crack extension is determined by the effective confining pressure (normal compression stress minus liquid pressure) if liquid pressure is constant. Although the validity of this assumption may depend on the liquid volume fraction and the fracture mode (opening or shear), this result shows that the presence of the melt phase can play a significant role in the intergranular cracking through its effect on the effective confining pressure. We also found that the stress concentration at the crack tip is relaxed to a great extent by matter diffusion, in addition to the plastic deformation usually considered in the brittle fracture theory. The effect of diffusional relaxation should also be taken into account. Therefore, to investigate the intergranular cracking experimentally, it is essential to control the "time constant" of the change of differential stress, which has not been paid much attention to in the previous deformation experiments on partially molten rocks.
Based on these understandings, we are now preparing experiments to investigate the effect of melt on the intergranular cracking using a partially molten rock analogue. In the presentation, we would also like to talk about the plan of these experiments.
In this study, we theoretically considered the condition for the intergranular cracking in partially molten rock. First, we examined the applicability of the Navier-Coulomb criterion, which states that the shear stress needed for fracture is proportional to the normal compressive stress. This criterion is explained by the increase in the real contact area in proportion to the normal compressive stress applied to a rough surface. Since such a rough surface does not exist in the partially molten rocks, we concluded that the Navier-Coulomb criterion is not applicable. Therefore, the effective confining pressure theory based on the Navier-Coulomb criterion is not applicable, either. Next, we applied the Griffith theory to the crack extension at the junction of grain boundary and two solid-liquid interfaces in the partially molten system. We found that the condition for the crack extension is determined by the effective confining pressure (normal compression stress minus liquid pressure) if liquid pressure is constant. Although the validity of this assumption may depend on the liquid volume fraction and the fracture mode (opening or shear), this result shows that the presence of the melt phase can play a significant role in the intergranular cracking through its effect on the effective confining pressure. We also found that the stress concentration at the crack tip is relaxed to a great extent by matter diffusion, in addition to the plastic deformation usually considered in the brittle fracture theory. The effect of diffusional relaxation should also be taken into account. Therefore, to investigate the intergranular cracking experimentally, it is essential to control the "time constant" of the change of differential stress, which has not been paid much attention to in the previous deformation experiments on partially molten rocks.
Based on these understandings, we are now preparing experiments to investigate the effect of melt on the intergranular cracking using a partially molten rock analogue. In the presentation, we would also like to talk about the plan of these experiments.