Smoothing of Fault Slip Surfaces by Scale‐Invariant Wear

Abstract

AbstractPrevious field observations suggest that fault slip surface roughness may decrease with slip. However, measurements have yet to confidently isolate the effect of slip from other possible controls, such as lithology or tectonic setting. We describe the evolution of slip surfaces in normal faults in SE Utah that cut well‐sorted, high‐porosity sandstones and accommodated regional extension at 1‐ to 4‐km depth. Tight controls on fault offset and uniform tectonic history allowed us to isolate the effect of slip during early stages of faulting (0‐ to 55‐m slip). Slip surfaces progress from rough joints and deformation bands toward smooth, continuous, and mirror‐like surfaces with increasing slip. We collected 123 scans of pristine slip surfaces, which measure surface geometry from micrometers to meters in length scale. Results indicate that slip surface roughness is systematically smoother than joint surfaces and deformation band edges. Over the best resolved length scales (0.1–10 mm), we observe roughness decreases with slip according to a power law with exponent −1 in the slip direction. Slip‐perpendicular profiles, though rougher, exhibit the same smoothing trend. The Hurst scaling exponent does not change with slip. These observations require wear to be multiscale. Boundary element method models suggest that mechanical wear of completely mated surfaces occurs by asperity failure and that the wear rate depends on the aspect ratio of asperities. These results indicate that at large slip, asperity failure at all length scales can cause slip surfaces to smooth while maintaining the fractal geometry characteristic of faults.