10:45 AM - 11:00 AM
[SCG62-01] Image analysis of dynamically recrystallized quartz from high-PT deformation experiments: Temperature and strain-rate effects on grain shapes
Keywords:quartz, deformation experiments, recrystallization, dislocation creep, image analysis
Recrystallization microstructures of quartz in high-pressure metamorphic belts and ductile shear zones vary with the metamorphic grades, suggesting that microstructural evolution reflects deformation conditions such as temperature and strain rates. To investigate this relationship, we conducted high-pressure (P) and temperature (T) deformation experiments and microstructure observations.
Masuda and Fujimura (1981, Tectonophysics) (M&F) deformed agate at 0.4 GPa using a solid-medium apparatus (Kumazawa apparatus) and identified two types of steady-state microstructures: Type S with oblate grains and serrated boundaries at low-T/high strain rate conditions, and Type P with polygonal grains at high-T/low strain rate ones. In contrast, Hirth and Tullis (1992, JSG) (H&T) deformed natural quartzite at 1.5 GPa using a Griggs apparatus and classified deformation and recrystallization microstructures into three regimes. The run products obtained at low-T/high-strain rate conditions (regime 3) resemble Type P of M&F. However, at low-T/high-strain rate conditions, original coarse grains remained in most samples, and recrystallized grain fractions were far smaller than 100%. Thus, whether or not quartz microstructures approach Type S and P of M&F at the steady states under higher confining pressure remains unsolved.
To address pressure dependence of steady-state quartz microstructures, we conducted deformation experiments on agate at 1.5 GPa confining pressure at the temperature and strain-rate conditions (800-1000 deg C, 10-4-10-6 sec-1) that largely cover the previous experiments at 0.4 GPa by M&F, using a new Kumazawa-type apparatus, which enables accurate differential stress measurement via a couple of load cells attached to the upper and lower pistons (Shimizu & Michibayashi, 2022, Minerals). Optical microscopic observations revealed quartz microstructures similar to Type S and P.
For quantitative analysis of grain shapes, we prepared six microphotographs taken under different polar angles. Using an image processing software, the grain boundaries were semi-automatically identified with a gradient filter. We then calculated normalized perimeter PN (perimeter/circumference of an equivalent-area circle), aspect ratio R, and perimeter-area fractal dimension D. Whereas the systematic change of D with strain rate as reported by Takahashi et al. (1998, JSG) was not clearly observed, PN and R monotonically increased from high-T/low-strain rate to low-T/high-strain rate conditions, which enables to capture the Type S/P transition. The optimum S/P boundary values were chosen as PN~1.37 and R~2.30. To further assess the effect of pressure, we conducted an additional experiment at 900 deg C and 10-5 sec-1 under lower confining pressure of 0.4 GPa; M&F obtained a Type P quartz microstructure in the same condition. Our 0.4 GPa sample showed PN = 2.05 and R = 1.40 of Type S. However, compared to the 1.5 GPa run product, which yielded PN = 2.48 and R = 1.45 (Type S), these values were closer to Type P. It is suggested that the S/P boundary shifted slightly toward higher-T/lower-strain rate conditions at higher pressure. A possible reason for this shift is the changes in dominant slip systems with increasing confining pressures.
Masuda and Fujimura (1981, Tectonophysics) (M&F) deformed agate at 0.4 GPa using a solid-medium apparatus (Kumazawa apparatus) and identified two types of steady-state microstructures: Type S with oblate grains and serrated boundaries at low-T/high strain rate conditions, and Type P with polygonal grains at high-T/low strain rate ones. In contrast, Hirth and Tullis (1992, JSG) (H&T) deformed natural quartzite at 1.5 GPa using a Griggs apparatus and classified deformation and recrystallization microstructures into three regimes. The run products obtained at low-T/high-strain rate conditions (regime 3) resemble Type P of M&F. However, at low-T/high-strain rate conditions, original coarse grains remained in most samples, and recrystallized grain fractions were far smaller than 100%. Thus, whether or not quartz microstructures approach Type S and P of M&F at the steady states under higher confining pressure remains unsolved.
To address pressure dependence of steady-state quartz microstructures, we conducted deformation experiments on agate at 1.5 GPa confining pressure at the temperature and strain-rate conditions (800-1000 deg C, 10-4-10-6 sec-1) that largely cover the previous experiments at 0.4 GPa by M&F, using a new Kumazawa-type apparatus, which enables accurate differential stress measurement via a couple of load cells attached to the upper and lower pistons (Shimizu & Michibayashi, 2022, Minerals). Optical microscopic observations revealed quartz microstructures similar to Type S and P.
For quantitative analysis of grain shapes, we prepared six microphotographs taken under different polar angles. Using an image processing software, the grain boundaries were semi-automatically identified with a gradient filter. We then calculated normalized perimeter PN (perimeter/circumference of an equivalent-area circle), aspect ratio R, and perimeter-area fractal dimension D. Whereas the systematic change of D with strain rate as reported by Takahashi et al. (1998, JSG) was not clearly observed, PN and R monotonically increased from high-T/low-strain rate to low-T/high-strain rate conditions, which enables to capture the Type S/P transition. The optimum S/P boundary values were chosen as PN~1.37 and R~2.30. To further assess the effect of pressure, we conducted an additional experiment at 900 deg C and 10-5 sec-1 under lower confining pressure of 0.4 GPa; M&F obtained a Type P quartz microstructure in the same condition. Our 0.4 GPa sample showed PN = 2.05 and R = 1.40 of Type S. However, compared to the 1.5 GPa run product, which yielded PN = 2.48 and R = 1.45 (Type S), these values were closer to Type P. It is suggested that the S/P boundary shifted slightly toward higher-T/lower-strain rate conditions at higher pressure. A possible reason for this shift is the changes in dominant slip systems with increasing confining pressures.