The work of fault growth in laboratory sandbox experiments

Contractional sandbox experiments that simulate crustal accretion and direct shear tests both provide direct data on the amount of work required to create faults (Wprop) in granular materials. Measurements of force changes associated with faulting reveal the work consumed by fault growth, which can be used to predict fault growth path and timing. Within the contractional experiments, the sequence and style of early faulting is consistent for the range of sand pack thicknesses tested, from 12 to 30 mm. Contrary to expectations that Wprop is only a material property, the experimental data show that for the same material, Wprop increases with sand pack thickness. This normal stress dependence stems from the frictional nature of granular materials. With the same static and sliding friction values, incipient faults initiated deeper in the sand pack have larger shear stress drops, due to increased normal compression, σn. For CV32 sand, the relationship between Wprop and σn, calculated from the force drop data as Wprop (J/m2)=2.0×10−4 (m) σn (Pa), is consistent with the relationship calculated from direct shear test data as Wprop (J/m2)=2.4×10−4 (m)σn (Pa). Testing of different materials within the contractional sandbox (fine sand and glass beads) shows the sensitivity of Wprop to material properties. Both material properties and normal stress should be considered in calculations of the work consumed by fault growth in both analog experiments and crustal fault systems.