The Role of Fabric in Frictional Properties of Phyllosilicate-Rich Tectonic Faults

The overall goal of this procedure is to test the frictional properties of phyllosilicates, with faults sheared in the in-situ geometry, and to show that this friction is significantly lower than friction of powders obtained by the same material. During the long-term evolution of tectonic faults, numerous geological studies have documented fluid-assisted reaction softening, that promotes the replacement of strong and granular minerals with phyllosilicates. In particular, fracturing processes along faults increase permeability, and facilitate the influx of hydrous fluids into the fault zone.

Fluids react with fine-grained rock, promoting dissolution of the strong minerals like quartz, feldspar and calcite. They become platy phyllosilicates and form foliated microstructures, like the one presented here in green. Slip along the phyllosilicates from the micro-scale is transmitted to the entire fault zone via the interconnectivity of the phyllosilicate-rich shear zones.

This is an example of the continuity of the phyllosilicate shear zone at the outcrop scale, that can be extended up to crustal-scale faults with thicknesses of more than 100 meters. Along phyllosilicate-rich fault like this one, the tonic shearing has produced the phyllosilicate alignment, producing this fault rock anisotropy. In order to take into account the role of anisotropy in frictional properties of the fault, we have to collect the right rock samples.

To do that, we have to collect a representative rock sample, and within the outcrop, we select a portion where the kinematic indicators are best preserved. And then we use a chisel and a hammer to collect the rock sample. Once the rock sample has been collected, we mark the sense of shear, and then we bring the rock sample to the lab for the experiment.

With this procedure, we cut the rock samples to obtain wafers that fit the forcing blocks of the rock deformation apparatus. This is usually achieved in 2 steps. In the first step, we use a standard laboratory saw to obtain rock samples that are slightly larger than the forcing blocks.

Secondly, we use a high precision rotary blade, or a hand grinder, to shape the wafers so they're 5 by 5 centimeters in area, and about 1 centimeter in thickness. From the same piece of rock, we use a disk mill to obtain a granular material that is sieved to reach the desired grain size, usually below 125 microns. The 2 identical wafers are mounted on stainless steel forcing blocks with a nominal frictional area of contact of 5 by 5 centimeters, and then are assembled with a central forcing block to compose the symmetric, double-direct configuration.

In the same way, the powders are used to construct 2 identical layers whose thickness is about 5 millimeters, and whose area of contact is 5 by 5 centimeters. These are then used to compose a similar double-direct shear configuration. At this point, the double-direct shear configuration is positioned within our biaxial apparatus, and we're ready to start the friction experiment.

We use a servo-controlled hydraulic piston to apply and maintain a constant normal stress on the rock sample. Then by advancing the vertical ram, we apply shear stress at constant sliding velocity;it is usually 10 microns per second. Most of the experiments are characterized by an initial strain hardening, where the shear stress increases rapidly during elastic loading, followed by shear stress at steady state.

The shear stress to normal stress ratio gives us the friction coefficient. At the end of the friction test, we carefully extract the experimental fault, we impregnate the rock sample with epoxy resin, we cut the sample in a direction parallel to the sense of shear, and we build thin sections from the cuts for microstructural studies. We use an optical microscope to characterize the bulk faults on the microstructure.

We analyze microstructures with a scanning electron microscope to investigate the main deformation processes. We use a transmission electron microscope to obtain details about deformation processes down to the nanoscale. In a diagram of normal stress versus shear stress, both the solid foliated wafers and the powder samples plot along the line, consistent with a brittle failure envelope.

But the solid wafers have a friction value significantly lower than the powdered analogs. In particular, the powders show friction of about 0.6, whereas the foliated rocks have significantly lower values. At each normal stress, the foliated rocks have a friction coefficient that is 0.2 to 0.3 units lower than the powders made from them.

Microstructural studies of the tested rocks show that the low friction of the solid wafers is due to sliding along the pre-existing, very fine-grained foliations made of phyllosilicates. TEM images show that slip is mainly accommodated by fracturing, translation, and rotation along the phyllosilicates, with frequent inner layer delamination. In contrast, experiments conducted on powders indicate that much of the deformation occurs along zones effected by fracturing and grain size reduction.

This results in higher values of friction. This is a summary of the frictional properties of natural, phyllosilicate-rich tectonic faults from different tectonic environments. Data show that friction is in the range of 0.1 to 0.3, and this friction is significantly lower than the traditional Byerlee value of friction obtained from a large gamut of rock types, that are predominantly made of granular mineral phases.

To summarize, our friction experiments show that foliated samples are extremely weak compared to their powdered equivalents. Microstructural studies indicate that the lower friction, or in other words, fault weakness of the foliated fault rocks is due to the reactivation of the pre-existing natural phyllosilicate-rich surfaces. These surfaces are absent in the powdered samples since the sample preparation step destroys them.

Our friction tests on solid foliated samples show that low friction, and therefore fault weakness, can occur in cases where weak mineral phases constitute only a small percentage of the total fault rock, implying that a significant number of crustal faults are weak.