The Role of Stress Field on Fault Reactivation: What We Can Learn from Experimental Rock Deformation

Fault mechanics predicts that fault reactivation and slip occur when the shear stress exceeds the fault strength, potentially nucleating earthquakes. Following Anderson’s legacy, the evolution of tectonic stress over the seismic cycle results in different couplings between normal and shear stress. This coupling depends on (1) the fault’s orientation relative to the maximum principal stress and (2) the tectonic faulting style. Reverse faults, loaded by an increase in maximum principal stress, experience increases in both normal stress and shear stress during the interseismic phase. In contrast, normal faults, loaded by a decrease in minimum principal stress, undergo a reduction in normal stress as shear stress builds up. For instance, low-angle normal faults experience larger increases in normal stress for the same shear stress increment compared to Andersonian 60°-dipping normal faults.

Despite this rich variety of stress field evolution observed in nature, laboratory deformation experiments have predominantly focused on a single stress-field scenario: a reverse fault optimally oriented for reactivation. The choice of reversed faults is dictated by the geometry of the apparatus, and the optimal orientation is the simplest system to generate recurrent lab-quakes. This simple laboratory approach describes faults as planes embedded in elastic media.

Here, I summarize results we have collected in the past years by systematically investigating in the laboratory the role of fault orientation and tectonic faulting style under triaxial saw-cut configuration—broadening the range of scenarios beyond the single one described above. The results reveal the impact of stress field on both fault zone and surrounding host rock deformation.

For fault zones, our results on gouge-bearing faults show clear discrepancies when compared with theoretical reactivation based on Coulomb-Mohr criterion. Faults at higher angles to the maximum principal stress appear weaker, suggesting potential stress field rotation within the fault zone. Additionally, when the normal stress for reactivation is comparable, reverse faulting tends to promote stable creep, while normal faulting—due to greater compaction and stiffness of the fault zone—favors slip acceleration and instabilities.

Fault orientation also affects the stress state of the surrounding host rock over the seismic cycle. Optimally oriented faults behave like ideal spring-slider system: elastic energy accumulates in the host rock during the interseismic phase and is released via on-fault slip during the co-seismic phase, accompanied by precursor acoustic activity. In contrast, unfavorably oriented faults produce a more complex picture. The host rock becomes critically stressed, acoustic activity spreads throughout the host rock, and precursors to lab-quakes become undetectable.

These results highlight the potential of investigating in the laboratory the role of stress field on fault zone deformation and its interplay with the surrounding host rock during earthquake nucleation. By expanding laboratory observations to include a wider range of stress-field scenarios, we take one small step toward bridging the gap between simplified experiments and the complex fault systems observed in nature.