Frictional Behavior Of Simulated Biotite-bearing Fault Gouge Under Hydrothermal Conditions

The cause of weakness of major crustal fault zones has been a long-livedresearch focus. Investigation of the mechanical properties of weak minerals throughlaboratory experiments offers a way to probe the weak domains. Biotite is a commonphyllosilicate in the nature and it has a low frictional strength from previous studies atroom temperature. Results of melting experiments show that biotite is chemicallystable in a very wide temperature range, up to700°C at least. As biotite is chemicallystatble in the seismogenic zone in the crust and it is often seen in fault rocks,investigation on its frictional sliding at hydrothermal conditions may help understandthe origin of weakness, the mechanical behavior in general and earthquake genesis inweak fault zones. Due to its low strength and layer structure, boitite tends to formfoliations and interconnected networks in fault zones. Thus, on the basis ofexperimental work on pure biotite gouge layer, further investigation of the effect oflayered strcture and foliations associated with biotite on fault sliding behavior(fictional strength and stability) is an equally important topic.Based on the background mentioned above, we performed suites of shearingexperiments on biotite gouge and biotite-quartz mixture gouge from room temperatureto elevated temperatures. Major attention is paid here to both frictional strength andrate dependence of friction.For the biotite gouge, the coefficient of friction changes with temperature in arange of~0.25to~0.44. Similar to muscovite, the coefficient of friction shows anincreasing trend with temperature elevation, but it is significantly smaller than that ofmuscovite. The steady-state rate dependence shows velocity strengthening attemperatures from25°C to200°C, and transitions to velocity weakening attemperatures above300°C. Stick-slip motions occurred at temperatures from400°C to600°C, with quite minor (a-b) ranging from-0.0004to-0.0008and submicron dcvalues (0.3~0.8μm). Microstructures of deformed samples show three typescorresponding respectively to three different temperature ranges. While brittlestructures are dominantly associated with deformation at100°C, significant plasticdeformation occurred at higher temperatures. Deformation at temperatures of200°Cto400°C are characterized with shear fracture zones accompanied with plasticdeformation mostly associated with basal glide. Intensively deformed zones are thecharacteristic structures for samples deformed at500°C to600°C, accompanied withshear fracture zones and moderately deformed zones.The results of structured gouge show that a layer in the fault gouge can lead tosignificant weakness in the fault zone. At100°C, the strength of homogeneously-mixed gouges shows a linear decreasing trend with increasing biotite proportion.However, for structured gouges with biotite layers embedded in quartz gouge, thestrength has a power-law decreasing trend with increasing weight proportions of biotite. The fault gouges can be weakened significantly by as little as~5wt%biotite,and~30wt%biotite corresponds to a beginning point of less sensitive strength changein response to increasing biotite proportion. When the structured gouge and mixturegouge both contain~19%wt biotite, they have the largest strength difference.Results in this study at elevated temperatures also show that the layered structurein the fault gouge play an important part in both strength and sliding stability of afault zone. The friction coefficient of the structured gouge changes with temperaturein a range of~0.4to~0.6as temperature elevates from25°C to600°C, a littlestronger than that of biotite gouge by35%. Moreover, both the structured gouge andbiotite gouge show an increasing trend with temperature elevation. At effectivenormal stress of200MPa, the steady-state rate dependence of structured gouge show atransition from velocity strengthening to velocity weakening around200°C. Similar tothe biotite gouge, stick-slip motions occurred at temperatures from400°C to600°CFor the structured sample, with quite minor|a-b|around average value of0.0007andsubmicron dcvalues (average value:~0.4μm). At a higher effective normal stress of230MPa and temperature of600°C, the sliding behavior of the structured gougeevolved with sliding distance to viscous flow when the total displacement comes to avalue around~2.3mm.For the biotite-quartz mixed gouge, coefficient of friction varies ranges from0.6to0.7as temperatures elevates from25°C to600°C, showing less sensitivity totemperature change than the structured sample. The mixture gouges show differentsteady-state rate dependence with different sliding velocities. While velocityweakening only occurred at200°C under moderate velocity steps (between1.22and0.244μm/s),,the steady-state rate dependence for slow velocity steps (between0.244and0.0488μm/s) cases show velocity weakening at200°C,400°C and600°C.The microstructures of the deformed structured gouges at effective normal stressof200MPa show that the shear deformation mainly occurred in the weak biotitelayers with a number of shear bands including Riedel shears (R1), P shears, Y shearsand Boundary shears (B) pervasively developed. The quartz layers on two sides aremore and more compacted as temperature increases. Moreover, at higher temperaturesof500°C and600°C, R1shears are also observed in the quartz layers in somelocalities. The microstructure related the viscous flow behavior of the structuredgouge as deformed at effective normal stress of230MPa shows that the shear bandsnot only occur in the biotite layer, but also extend to the quartz layers. This structureindicates that the viscous flow is closely related to the shear deformation in quartz.For biotite-quartz mixture gouges, the deformation microstructures show extensivedevelopment of shear bands, with prevalent R1and B shears but seldom P shears andY shears. Biotite and quartz in mixed gouges are both sheared with boitite arranged inP orientation, and their grain sizes were reduced extremely as temperature increases.Based on results of biotite gouge, the low coefficients of friction of biotite gouge found in our experiments fails to explain the large-scale weakness of San AndreasFault, thus super-hydrostatic fluid pressures are also required for facilitating theapparent weakness. When the fluid overpressure factor (pore pressure/hydrostaticpressure) is~2.2, biotite in fault rocks with abundance sufficient to facilitateconnectivity and a network may produce unstable fault slips in the mid crust.The results of the structured gouge with biotite as layers embedded in quartzgouge are probably more similar to natural fault zones. The fault strength inferredfrom our experimental data of structured gouge is similar to that inferred from biotitegouge. Thus, similar to the biotite gouge, the low coefficients of friction found forstructured gouge in this study do not explain the apparent weakness of San AndreasFault as well, where super-hydrostatic fluid pressures are required accordingly.Nevertheless, the fault strength inferred from the structured gouge should be moresimilar to a fault zone with weakness than that inferred from the mixed gouge. Withbiotite within fault rocks forming foliations and interconnected networks, wespeculated that the seismogenic zone may extend to depth of mid-crust.