The Strength of Common Fault Gouge Minerals
A longstanding question in earthquake research is why the strength of the San Andreas Fault in California (the shearing force required before slip occurs) is relatively weak compared to the frictional strength of typical rocks as measured in the laboratory. One possibility is that the finely-ground and altered material within the fault zone (fault gouge) is inherently weak and allows the fault to slip more easily. Another possibility is that water can become trapped within the fault zone and lower its strength, essentially "lubricating" the fault. We find that although many fault gouge minerals are indeed weak, they are not weak enough to completely explain the low strength of the San Andreas Fault. However, high fluid pressures acting in combination with these weak gouge minerals can easily lead to fault rupture.
Fault Gouge Primer
What does fault gouge look like? As the rough surfaces of rock on either side of a fault slide past one another, the crushing and grinding process produces a fine, gritty debris known as "fault gouge". Over time, this crushed rock can react with subsurface fluids to produce a variety of other secondary minerals, many of them in the "clay" family. Often, fault gouge is a mixture of crushed rock and several of these fine-grained alteration minerals. However, some fault gouge may be composed of finely-ground particles of just one principle type of mineral. The "gouge zone", where the grinding and shearing takes place, may be up to a kilometer wide in large faults.
Many typical gouge minerals, such as those with sheeted crystal structures (like the leaves of a book), are characterized by low frictional strength (coefficient of friction, µ, <0.5) when saturated with water. A thin layer of water, a few molecules thick, is often held by the electrical surface charges of the mineral grains. This sheath of water can lubricate the mineral grains, making it easier for them to slip past each other. However, if this water is driven off by heat or pressure, some gouge minerals become strong like the majority of typical materials, with µ = 0.6-0.85 (Byerlee's Law). This variable strength behavior is obviously important for understanding the factors that determine fault strength. Consequently, we have studied the frictional properties of many types of granular gouge minerals under both dry and wet conditions to assess the effect of adsorbed (or interlayer) water on fault gouge strength.
The minerals that we have studied fall in several different categories mainly controlled by their crystal structure: the 3-dimensional framework-structure minerals calcite, albite, quartz and the zeolites laumontite and clinoptilolite; the serpentinites antigorite, lizardite and chrysotile; and the 2-dimensional (leaf-like), sheeted-structure minerals kaolinite, muscovite, chlorite, brucite, montmorillonite, talc, and graphite. A thin layer of finely-crushed gouge was placed between the two halves of a cylindrical rock containing a 30° sawcut, as shown in the "triaxial experiment" diagram. This rock-gouge-rock sandwich, representing a fault containing a layer of fault gouge, was dried for several days to drive off all internal water. Triaxial sliding experiments were then conducted at 100 MPa normal stress (the compressional stress clamping the fault surface closed) with a sliding velocity (axial shortening rate) of 0.5 mm/sec. The dry samples were slid for 4 mm of displacement, then saturated with water and slid an additional 5 mm displacement.
The strength of the various single-mineral gouges is plotted in terms of the coefficient of friction, µ, versus sliding displacement in Figure 1.
Figure 1. Coefficient of friction, µ, versus axial sliding displacement for fault gouge minerals. The samples were saturated at 4 mm displacement. (top left) Calcite, albite, and quartz; (top right) Zeolites: laumontite and clinoptilolite; (bottom left) Serpentines: antigorite, lizardite, and chrysotile; (bottom right) Sheet-structure minerals: kaolinite, muscovite, chlorite, brucite, montmorillonite, talc, and graphite.
After an initial elastic response during the first roughly 1 mm of displacement (steep slope), the dry strength of the gouges can be determined. The framework-structure minerals (calcite, albite, quartz, zeolites) are relatively strong, following Byerlee's Law (µ = 0.6-0.85). However, the sheeted-structure minerals show a broad range of strengths, with µ from 0.15 to 0.8. This is because the frictional strength of these minerals depends on the bond strength, and hence the type of chemical bond, between the layers. For instance, a dipole-dipole type bond (OH-OH) found in brucite is easier to separate than a hydrogen bond (OH-O) typical of kaolinite. Hence brucite (µ = 0.45) is weaker than kaolinite (µ= 0.85) when dry. This correlation is clearly seen if we plot dry friction against the energy required to separate layers of sheeted-structure minerals (Figure 2).
Figure 2. Maximum dry coefficient of friction, µ, (this study) versus separation energy of sheet-structure minerals, from Giese (1978, 1980) and Bish and Giese (1981).
Effect of Water
After the introduction of water into the samples at 4-mm
displacement, the framework-structure minerals showed little
or no change in frictional strength. Sheeted-structure minerals,
on the other hand, can be greatly weakened by adsorbed (or interlayer)
water because their charged surfaces readily attract water
molecules. The strength of these minerals is reduced by as
much as 65%, and many have a coefficient of friction, µ, as
low as 0.2. The tendency to adsorb water depends on many factors,
including the composition of the gouge and the dissolved minerals
in the water. As a group, the serpentine minerals (chrysotile,
lizardite, and antigorite) show the greatest tendency to be
weakened by water, followed by the sheeted-structure minerals
and then the zeolites (Figure 3).
Figure 3. Percentage decrease in frictional strength of gouge minerals after saturation
In spite of the weakening effect of adsorbed water, the gouge strength values are not sufficiently low enough to explain why faults such as the San Andreas in California are relatively weak. Heat flow and stress-drop measurements along the San Andreas Fault indicate that a coefficient of friction of less the 0.2 is required for slip. Apparently, some additional mechanism is required to produce slip on the San Andreas Fault. For example, abnormally high pore fluid pressures, acting alone or in combination with weak fault gouge minerals, could be responsible for the observed fault strength.
- The maximum (dry) coefficient of friction, µ, for sheeted-structure minerals correlates with the strength, and hence type, of interlayer bond.
- The coefficient of friction, µ, of sheeted-structure minerals is reduced by various degrees in the presence of adsorbed or interlayer water. The tendency of a mineral to adsorb water depends on many factors including composition, surface charge, chemistry and pH of the water.
- The combination of the above explains why many common fault gouge minerals are anomalously weak. However, pressure and temperature at seismogenic depth tends to increase the strength of fault gouge minerals by driving off adsorbed or interlayer water, such that gouge strength alone can not explain the apparent weakness of strike-slip faults such as the San Andreas.
Details of this study can be found in:
Morrow, C.A., Moore, D.E., and Lockner, D.A., The effect of mineral bond strength and adsorbed water on fault gouge frictional strength, Geophysical Research Letters, vol. 27, no. 6, 815-818, 2000.