On the role of phyllosilicates on fault lubrication: Insight from micro‐ and nanostructural investigations on talc friction experiments

Abstract

[1] Exposures of mature faults at the Earth's surface show that in the shallow crust some of co-seismic slip occurs within a narrow phyllosilicate-rich gouge material. Mechanical behavior of phyllosilicates sheared at seismic slip-velocities remains a complex issue as suggested by the experimental and theoretical studies reported so far. Talc represents a simple case for common phyllosilicates in natural faults and therefore we chose talc as an analogue to better understand the mechanical behavior of clay-rich mature crustal faults. We present a series of high-velocity friction experiments conducted on talc at 1.31 m s À1and normal stresses of 0.3-1.8 MPa for wet and dry conditions. At 1 MPa normal stress, both wet and dry experiments show a slip-weakening behavior, however, with a higher decreasing rate in wet conditions. Dynamic shear stress evolves from a peak value of 0.24-0.52 MPa down to a residual state value of 0.08-0.18 MPa in wet conditions, and from a peak value of 0.46-1.17 MPa down to a residual state value of 0.12-0.62 MPa in dry conditions. Based on a detailed microstructural analysis down to the nanoscale, we propose thermal-pressurization as a possible slip-weakening mechanism for wet conditions. On the contrary, in dry conditions the long-lasting weakening is interpreted to be due to a combination of i) progressive disappearance of geometrical incompatibilities, ii) solid lubrication of talc lamellae, and iii) powder lubrication by nanometric aggregates. We conclude that initial gouge humid conditions and inherited fabric from past sliding may have large influence on the slip-weakening for subsequent slip.

The frictional strength of talc gouge in high‐velocity shear experiments

“…Boutareaud et al [] conducted an in‐depth investigation of talc frictional properties. They sheared room‐dry and wet talc gouge at constant equivalent slip velocity of 1.31 m/s, under σ n = 0.3–1.8 MPa, and to slip distances of 2.99 m to 49.19 m. We noted several similarities between our results and those of Boutareaud et al [], including (1) similar static friction coefficient of ~0.3 for wet talc and ~0.8 for dry talc (Figure and their Figure 2); (2) similar trends (not magnitudes) of compaction in wet talc and dilation of dry talc (Figure and their Figure 2); (3) development of secondary shear surfaces in dry‐talc gouge that we termed R and P shears (Figures h–j) and they termed “C, S, and C′‐oriented shear planes” (their Figure 11d); and (4) similar trends of grain size changes: less intense pulverization in wet talc that Boutareaud et al [] noted as no‐change in clast‐to‐matrix ratio versus intense pulverization of dry talc (Figures d, h, and j and their Figure 12, right).…”

“…We observed systematic slip‐strengthening at all tested velocities (Figure ), and more than 3 orders of magnitude of slip, in both constant velocity and stepping velocity, in which μ approaches 1.0 after 5–6 m of slip (Figures b and c). Boutareaud et al [], on the other hand, observed strong slip‐weakening at equivalent velocity of 1.31 m/s to steady state friction of μ ~ 0.2 (their Figure 2f). This difference stems from differences of experimental design.…”

“…These two groups of experiments, low‐slip velocity and high‐slip velocity, revealed different frictional behaviors of talc: slip‐strengthening in the µm/s range [ Moore and Lockner , , ; Collettini et al , ] and eventual slip‐weakening at 1.3 m/s [ Boutareaud et al , ]. Since slip velocity is a central controlling parameter of rock friction [ Reches and Lockner , ; Di Toro et al , ], we investigate here the frictional behavior and mechanisms of talc gouge between subseismic and seismic velocities, namely, 0.002 m/s up to 0.66 m/s.…”

“…Finally, low‐velocity experiments (<1 mm/s) indicate that while talc has a low‐friction coefficient, it displays slip‐strengthening and velocity‐strengthening behaviors [ Moore and Lockner , , ; Collettini et al , ]. On the other hand, the high‐velocity experiments (1.3 m/s in Boutareaud et al []) revealed intense slip‐weakening. Therefore, talc may be regarded as a conditionally stable gouge that could not nucleate earthquakes but could slip dynamically when accelerated to high velocity by a propagating rupture.…”

Frictional behavior of talc‐calcite mixtures

“…The 20% talc sample shows a YPR geometry [ Logan et al ., ] of the experimental gouge layer (e.g., Figure c) with talc lamellae concentrated along R 1 shear planes. This microstructure favors a deformation mechanism characterized by frictional sliding along the talc foliae (Figure b), typically observed in pure talc shear experiments [e.g., Boutareaud et al ., ; Misra et al ., ], promoting a decrease in the frictional strength to values of μ ss ranging between 0.33 and 0.20. Similarly, Moore and Lockner [] in quartz/talc mixtures observed a marked weakening resulting from the formation of R 1 planes that cross the entire sample.…”