Acoustic Emissions and Seismic Tomography of Sandstone under Uniaxial Compression: Implications for the Progressive Failure in Pillars

Guang Zhang,Shihuai Zhang,Pei Guo,Shunchuan Wu
DOI: https://doi.org/10.1007/s00603-022-03155-5
IF: 6.518
2022-01-01
Rock Mechanics and Rock Engineering
Abstract:This work presents the spatio-temporal variations of microcracking and damage accumulation during the progressive failure in a low-porosity sandstone under uniaxial compression, using active ultrasonic surveys and passive acoustic emission monitoring. The combination of both techniques allows us to perform a joint inversion of source locations and P-wave velocity tomography of the specimen during the failure process. Based on the P-wave first arrivals from both the active surveys and numerous passive AE events, three-dimensional (3D) P-wave velocity structure and its temporal variations can be obtained in high resolution in spite of the limited active ray coverage. The reconstructed P-wave velocity structure indicates that the specimen becomes more heterogeneous due to damage accumulation. Within such a highly anisotropic medium, the ray paths are found to be curved, rather than straight as assumed in a homogeneous material. The curved ray paths are used to calculate the travel times of each AE event so as to obtain the accurate source location. We calculate the energy of the detected AE event and estimate the energy distribution. It shows that about 50% of the events account only for about 2.5% of the total energy while 1% of the events release 47.3% of the total energy. The events with large energy, albeit few, allow to trace the fractures. We also investigate the variations of the P-wave velocities along different characteristic ray paths during the failure process. Based on the joint inversion, the variations of the 3D P-wave velocity structure with deformation have been further investigated. It is found that the emergence of the low-velocity zone (LVZ) does not indicate the onset of fractures and that the LVZ is highly unpredictable in a spatio-temporal sense. The spatial correlation between the LVZ and AE events suggests that AE events usually appear on the edge of the LVZs, where there are high-velocity contrasts. This work suggests that the quantification of P-wave velocity changes has the potential for extracting timely precursory information of pillar bursts.
What problem does this paper attempt to address?