Weaker than quartz? – Strain localization mechanisms and rheology of fine-grained polymineralic rocks

The strength and deformation behaviour of the Earth are necessary parameters to model and fundamentally understand crustal scale deformation. In the case of the Earth’s continental middle crust, so far, most models use extrapolated physical parameters from monomineralic deformation experiments, assuming pure quartz to represent the weakest rheological phase at mid-crustal conditions. This is an oversimplification as the Earth’s continental middle crust mostly consists of polymineralic granitoid rocks, the rheology of which is unknown so far.

We present the first experimental study investigating the viscous rheology of a natural, fine-grained, granitoid rock. To unravel the complex deformational behaviour, it is crucial to combine an in-depth microstructural analysis with thorough estimations of rheological parameters. Cylindrical natural granitoid ultramylonite samples, composed of qtz + ab + K-fsp + bt + ep, with grain sizes of 125-15 μm are deformed in a Griggs type apparatus at T=650°C, confining pressure=1.2 GPa, strain rates=10^-3 to 10^-5s^-1, and 0.2 wt% H₂O added. Mechanical data are combined with light microscope, SEM, TEM, and quantitative image analysis to connect microstructures with stress and strain evolution.

Only through this combination we can show that grain size sensitive deformation processes, namely dissolution precipitation creep (DPC) lead to extreme grain size reduction, and pinning prevents grain growth and produces a stable microstructure. These processes lead strain localization and overall very weak viscous behaviour – weaker than can be extrapolated from monomineralic quartz. We further can show through two different experimental setups how strain is localizing with and without preexisting fracture in an initially foliated rock. Verified with microstructural observations, we fit parameters into a constitutive equation for this DPC, based on an exponential diffusion creep flow law, to model our experiments and tackle the extrapolation to various natural conditions.

Our findings imply that the Earth’s granitic middle crust deforms faster in localized shear zones than previously modelled. This would result in faster stress buildup at the viscous to brittle transition, promoting seismic ruptures in the overlaying brittle crust than predicted so far. Thereby, our new insights can improve models investigating mechanics of extensions of earthquakes to the middle crust, where they supposedly nucleate and help understand seismic cycles or improve models of stress buildups and thermal flow below and influences on hydro/geothermal systems. We further highlight the importance to improve modelling of polymineralic systems.