- 1Earth and Planetary Sciences Institute, School of GeoSciences, University of Edinburgh, Edinburgh, UK
- 2Lyell Centre Geoenergy Group, Heriot-Watt University, Edinburgh, UK
- 3National Oceanography Centre, Southampton, UK
- 4School of Engineering, University of Manchester, Manchester, UK
- 5RWTH Aachen University, Aachen, Germany
- 6Diamond Light Source, Didcot (Harwell), UK
Seismic velocity at the near surface drops during ground motions due to remote earthquakes and can recover afterwards over decades. In the laboratory, seismic velocity of rock samples decreases after dynamic deformations (e.g., shaking) and gradually recovers towards the original level. These observations at different scales are referred to as slow dynamics in granular materials (e.g., rocks and concrete), but the underlying mechanisms remain debated.
We explore the physics behind seismic velocity transients during and after dynamic deformations using Stór Mjölnir — a triaxial pressure loading apparatus featuring two piezoelectric transducers mounted in the top and bottom pistons and an X-ray transparent aluminium pressure vessel that houses a cylindrical core sample of Clashach sandstone (10 mm in diameter and 25 mm in length).
We present the mechanical, acoustic, and X-ray microtomography results of two triaxial loading experiments, conducted at room temperature with a confining pressure of 20 MPa and a pore fluid pressure of 5 MPa. Both experiments involve first increasing the ram pressure at a constant strain rate of 1x10-5 s-1 until the onset of sample yielding, indicated by a deviation from the linear stress–strain curve. In the first experiment, we further hold the ram pressure constant and then abruptly reduce the pressure by 30 MPa before rapidly returning the pressure to the previous hold level; this perturbation is repeated for 32 cycles until catastrophic failure of the rock sample. In the second experiment, we apply the same cyclic loading protocol after sample yielding, except for the abrupt pressure drop of 150 MPa; the sample survives only two loading cycles before catastrophic failure. These cyclic loading protocols are designed to induce transient seismic velocity responses, which are monitored by active acoustic surveys acquired every 8 s and in-situ 3D X-ray tomography synthesised every 6 min at the beamline I12-JEEP, Diamond Light Source (Oxfordshire, UK).
We observe nearly linear relationships between the small stress perturbations (30 MPa) and corresponding seismic velocity changes, indicating minimal slow dynamics in the rock sample. In contrast, large stress perturbations (150 MPa) cause nonlinear velocity changes, although the recovery time scale is limited by the small size of the experimental sample. The time-resolved 3D X-ray volumes from both experiments show no resolvable transient structural changes in the rock samples, despite ongoing microfracture accumulation and pore enlargement driven by background creep until catastrophic failure. These results demonstrate that active seismic waves can detect nonlinear velocity transients in triaxial loading experiments, which likely originate from microstructures (e.g., grain contacts) below the X-ray imaging resolution (voxel edge length ~ 7.9 µm). These experiments also motivate further study on seismic velocity transients using our next-generation experimental apparatus that accommodates larger samples (18 mm in diameter and 45 mm in length) and six acoustic transducers. Ultimately, we aim to assess seismic velocity transients as a proxy for rocks’ susceptibility to small stress perturbations, which could provide a method to map the proximity to catastrophic failure and hence help mitigate induced seismicity associated with hydraulic fracturing.
How to cite: Li, X., Chandler, M., Cartwright-Taylor, A., Freitas, D., Mangriotis, M.-D., Woldemichael, B., Liptak, A., Atwood, R., Chapman, M., Fusseis, F., Butler, I., Curtis, A., and Main, I.: Exploring seismic velocity transients using in-situ acoustic monitoring and synchrotron-based X-ray tomography in triaxial dynamic loading experiments, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8295, https://doi.org/10.5194/egusphere-egu26-8295, 2026.