Fault rock heterogeneity can produce fault weakness and reduce fault stability

The study investigates how spatial heterogeneity in fault rock compositions influences fault strength and frictional stability. Fault zones commonly contain mixed materials, yet the mechanical implications of lateral heterogeneity are not well quantified. The authors test the hypothesis that introducing discrete patches of strong, rate-weakening quartz and weak, rate-strengthening clay into a fault gouge layer reduces overall strength and stability relative to a homogeneously mixed equivalent. Understanding these effects is important for explaining the spectrum of natural fault slip behaviours, from aseismic transients to earthquakes, and for improving models of earthquake nucleation and propagation.

Prior work often attributes the diversity of slip behaviours on natural faults to large-scale heterogeneities, while small-scale heterogeneities are also abundant in fault zones. Experimental studies on quartz gouges have shown that microstructural evolution and localization into discrete shear bands are prerequisites for unstable stick-slip. Stability analyses of rate-and-state faults indicate that slip instability requires sufficiently large rate-weakening patches. Numerical modelling suggests that decreasing the size of a rate-weakening patch stabilizes the response. These insights frame the present laboratory investigation of small-scale compositional heterogeneity in mixed clay-quartz gouges.

Laboratory friction experiments were conducted using a direct-shear arrangement within a triaxial deformation apparatus. Gouge layers (~1.3 mm initial thickness prior to pressurization) were prepared either as laterally heterogeneous patches (strong, rate-weakening quartz; weak, rate-strengthening clay) or as homogeneous mixtures of quartz and clay with identical bulk composition. The layers were placed between direct-shear forcing blocks. Soft silicone spacers were positioned at each end to accommodate displacement without supporting load. To discourage boundary shear, the sliding surfaces (50 × 20 mm area) on the forcing blocks were grooved perpendicular to the sliding direction (grooves 200 µm deep, 400 µm spacing). The assembled gouge layer was wrapped in a low-friction PTFE sleeve (0.25 mm thickness) to minimize jacket friction near the layer, then enclosed in a soft 3 mm thick PVC jacket (Nalgene 180 clear tubing). The jacketed direct-shear assembly was placed between platens and inserted into the pressure vessel. In this configuration, normal stress on the gouge was applied by confining pressure. Pore-fluid pressure was introduced through three porous disks embedded in each direct-shear block. Velocity-step tests were performed to evaluate the rate-and-state friction parameter (a−b) as a function of clay fraction and displacement, enabling comparisons between heterogeneous and homogeneous configurations.

• Spatial heterogeneity significantly reduces both strength and frictional stability relative to compositionally identical, homogeneously mixed gouges.
• Across clay fractions, heterogeneous faults consistently exhibit lower (a−b) values than homogeneous faults after 1.5 mm and 4.5 mm displacement, indicating reduced stability; the difference Δ(a−b) is predominantly negative.
• For displacements >1.5 mm, (a−b) could not be calculated for the homogeneous quartz endmember (0% clay) and the heterogeneous fault with 20% clay because velocity steps triggered stick-slip instabilities.
• Heterogeneous faults remain overall stable (positive (a−b)) when the proportion of rate-weakening material is ≤70%, though values are closer to zero (more rate-neutral) than in homogeneous counterparts.
• Stick-slip instabilities occur only when the strong rate-weakening patch comprises ≥80% of the layer.
• Identified mechanisms of weakening and destabilization include: (1) smearing of weak clay; (2) differential compaction of quartz vs. clay, redistributing normal stress; and (3) shear localization within strong quartz patches, producing stress concentrations.
• Decreasing the size or proportion of rate-weakening patches leads to a more stable response, consistent with rate-and-state stability criteria.

The experiments directly demonstrate that lateral compositional heterogeneity in fault gouges reduces strength and promotes less stable frictional behavior compared to homogeneous mixtures. This addresses the research question by showing that interactions between rate-weakening and rate-strengthening patches alter effective frictional parameters (lower (a−b)) and can trigger instabilities when the rate-weakening fraction is sufficiently large. The identified mechanisms—clay smearing, differential compaction with normal-stress redistribution, and shear localization causing stress concentrations—provide a physical basis for the observed weakening. These findings are relevant to natural faults where heterogeneity occurs across scales: smaller or fewer rate-weakening patches stabilize slip, whereas larger or more connected rate-weakening regions can nucleate unstable slip. The results bridge laboratory observations with theoretical and numerical expectations from rate-and-state friction and suggest that even small-scale heterogeneity can modulate the occurrence of slow slip versus earthquakes, with implications for earthquake source modeling and hazard assessment.

Introducing simple lateral heterogeneity into clay–quartz fault gouges substantially reduces fault strength and overall stability relative to homogeneous mixtures. The study identifies multiple mechanisms for weakening and destabilization and shows that instabilities arise when rate-weakening patches dominate (≥80%), whereas systems with ≤70% rate-weakening material remain rate-strengthening overall. Given the widespread occurrence of multiscale heterogeneity in natural faults, these interactions likely control whether faults slip aseismically or in earthquakes. The work underscores the need for additional laboratory experiments and earthquake source modeling to quantify the effects and evolution of small-scale heterogeneity and to incorporate these effects into larger-scale constitutive laws relevant for earthquake nucleation and induced seismicity.

• Precise publication data report mainly relative stability metrics ((a−b)) and qualitative mechanisms; at larger displacements (>1.5 mm) certain configurations (0% clay homogeneous; 20% clay heterogeneous) experienced stick-slip during velocity steps, preventing (a−b) determination.
• Instability requires a sufficient fraction of rate-weakening material; thus, findings on instability thresholds (e.g., ≥80%) pertain to the tested compositions and laboratory conditions.
• The experiments examine specific clay–quartz systems and small-scale heterogeneity; extrapolation to natural faults with more complex mineralogies and structures should be made cautiously.
• Small-scale heterogeneities may be difficult to represent in large-scale models, potentially limiting direct incorporation of these effects without further upscaling studies.