Rupture Velocity Dependence of Fracture Energy for Sub-Rayleigh Ruptures on a 6-Meter-Long Laboratory Fault

Kurama Okubo, Futoshi Yamashita, Yoshiaki Matsumoto, & Eiichi Fukuyama

Submitted September 7, 2025, SCEC Contribution #14680, 2025 SCEC Annual Meeting Poster #TBD

Fracture energy governs the earthquake rupture process from nucleation and propagation to arrest. Whether it can serve as an intrinsic constant associated with the physical state of the fault surface is of great interest for describing the rupture process and its scaling to natural earthquakes. In this study, we estimated the fracture energy by fitting a steady-state rupture model with a linear cohesive zone (Poliakov et al., 2002; Kammer and McLaskey, 2019) to local shear stress changes associated with rupture propagation recorded by a dense strain gauge array installed on a newly developed meter-scale biaxial rock friction apparatus.

The simulated laboratory fault was 6 m in length and 0.5 m in width. Strain gauges were mounted 15 mm away from the fault surface at 130 mm intervals along the side surfaces of the rock specimen. The experiment was conducted in two phases. First, we applied a normal stress of 3 MPa using six hydraulic jacks and carried out conventional stick–slip experiments. Second, we applied shear loading close to the peak frictional strength and then partially reduced the normal stress by controlling one normal loading jack to nucleate rupture at a predetermined location and timing. This allowed the rupture to propagate long enough to approach the limiting speed, i.e., the Rayleigh wave speed (c_R).

The steady-state rupture model incorporates rupture velocity, fracture energy, and cohesive zone size as input parameters, and once these are specified, it can describe the spatial distribution of the dynamic stress field near the rupture tip. Among these parameters, rupture velocity can be directly estimated. We measured it accurately using normalized cross‐correlation of shear stress between neighboring gauges, then fixed it to search for the fracture energy and cohesive zone size that best fit the measured shear stress time history.

After applying thresholds on the maximum correlation coefficient to ensure fitting quality, we collected fracture energy estimates for rupture velocities from 0.8 c_R to 0.99 c_R. The estimates varied from 0.07 to 0.005 J/m^2, decreasing as the rupture velocity approached c_R. These results provide experimental evidence that fracture energy inferred from the dynamic stress field can vary with rupture evolution, which may be consistent with the dynamic rupture theory of Freund (1990). Further investigation is required to clarify how this variation relates to an intrinsic property of the fault.

Key Words
fracture energy estimation, dynamic rupture, laboratory experiment

Citation
Okubo, K., Yamashita, F., Matsumoto, Y., & Fukuyama, E. (2025, 09). Rupture Velocity Dependence of Fracture Energy for Sub-Rayleigh Ruptures on a 6-Meter-Long Laboratory Fault. Poster Presentation at 2025 SCEC Annual Meeting.

Related Projects & Working Groups
Fault and Rupture Mechanics (FARM)