4:15 PM - 4:30 PM
[SCG44-04] Effect of dislocations on polycrystal anelasticity
Keywords:Anelasticity, dislocation, in-situ measurement, polycrystal
Rock anelasticity plays an essential role in extracting the upper mantle conditions such as temperature and grain size from seismic wave velocity and attenuation (e.g., Karato, 1993, Faul & Jackson, 2005; Priestley et al, 2024). Although there has been a large progress in understanding the polycrystal anelasticity caused by the grain boundary sliding mechanism (e.g., Gribb & Cooper, 1998; Jackson et al., 2002; Yamauchi & Takei, 2016), anelasticity caused by dislocations has been poorly understood. Although possible effects of dislocations on the dispersion and attenuation of seismic waves have been pointed (Anderson & Minster 1981), only a few experimental studies have been performed on the effect of dislocations on anelasticity (Gueguen et al., 1989: Faula et al., 2012; Sasaki et al., 2019). Moreover, in these previous studies, anelasticity was measured under low deviatoric stress by using the samples to which dislocations were introduced by the pre-deformation under high stress. In such experiments, the obtained anelasticity data could be affected by dislocation recovery and by insufficient coupling between sample and apparatus. In this study, we investigated the effect of dislocations on the polycrystal anelasticity by in-situ measurement of Young's modulus and attenuation under high deviatoric stress.
We used polycrystalline borneol as a rock analogue. Sasaki et al. (2019) reported that dominant deformation mechanism of this material at 50 C and at the grain size of about 20 μm changes from grain boundary diffusion creep to dislocation (power-law) creep at about 1.5 MPa. We developed a new experimental apparatus with two loading systems. The high deviatoric stress needed for the dislocation creep was applied by a commercial loading frame (Shimadzu Corporation, AGX-V 50kN). For the in-situ measurement of sample anelasticity, cyclic stress with a small peak-to-peak amplitude was generated by a multilayer piezoelectric actuator (TOKIN Corporation, ASB400C702WD1-A0LF) attached to the piston of the commercial loading frame. We measured stress applied to the sample by using the load cell which was placed within a triaxial cell used as a hot water bath to control the sample temperature. Confining stress was not applied in the present experiment. The oscillatory deformation of the sample was measured by using a pair of laser displacement meters (KEYENCE CORPORATION, CL-P015, resolution of 0.003 μm). The creep deformation of the sample was measured by the displacement meter of the loading frame.
Using this apparatus, a rock analogue sample was deformed by dislocation creep for about 19 hours and then deformed by diffusion creep for about 7 hours in succession. All through these deformations, the anelastic property of the sample was repeatedly and non-destructively monitored every 100 s by forced oscillation tests at a frequency of 5 Hz for a duration of 6 s. A remarkable result is that we did not detect any significant change of Young's modulus through the loading, dislocation creep, unloading, and diffusion creep. This result suggests that the effect of dislocations on the anelastic relaxations at >5 Hz is very small. The present result is different from the previous result by Sasaki et al (2019) that the dislocation-induced attenuation peak exists at higher frequency than 100 Hz, which decreases the Young’s modulus at 5 Hz by about 10%. The present result also shows that the confining stress is not needed to suppress micro cracking under the present experimental conditions.
We used polycrystalline borneol as a rock analogue. Sasaki et al. (2019) reported that dominant deformation mechanism of this material at 50 C and at the grain size of about 20 μm changes from grain boundary diffusion creep to dislocation (power-law) creep at about 1.5 MPa. We developed a new experimental apparatus with two loading systems. The high deviatoric stress needed for the dislocation creep was applied by a commercial loading frame (Shimadzu Corporation, AGX-V 50kN). For the in-situ measurement of sample anelasticity, cyclic stress with a small peak-to-peak amplitude was generated by a multilayer piezoelectric actuator (TOKIN Corporation, ASB400C702WD1-A0LF) attached to the piston of the commercial loading frame. We measured stress applied to the sample by using the load cell which was placed within a triaxial cell used as a hot water bath to control the sample temperature. Confining stress was not applied in the present experiment. The oscillatory deformation of the sample was measured by using a pair of laser displacement meters (KEYENCE CORPORATION, CL-P015, resolution of 0.003 μm). The creep deformation of the sample was measured by the displacement meter of the loading frame.
Using this apparatus, a rock analogue sample was deformed by dislocation creep for about 19 hours and then deformed by diffusion creep for about 7 hours in succession. All through these deformations, the anelastic property of the sample was repeatedly and non-destructively monitored every 100 s by forced oscillation tests at a frequency of 5 Hz for a duration of 6 s. A remarkable result is that we did not detect any significant change of Young's modulus through the loading, dislocation creep, unloading, and diffusion creep. This result suggests that the effect of dislocations on the anelastic relaxations at >5 Hz is very small. The present result is different from the previous result by Sasaki et al (2019) that the dislocation-induced attenuation peak exists at higher frequency than 100 Hz, which decreases the Young’s modulus at 5 Hz by about 10%. The present result also shows that the confining stress is not needed to suppress micro cracking under the present experimental conditions.