Triggering of rockburst in layered composite rocks

Abstract

Layered composite rocks are often encountered in deep underground engineering. Due to the high in-situ stresses and engineering disturbances, rockburst becomes one typical form of dynamic failures in layered composite rocks. However, investigations of rockbursts in layered composite rocks in the laboratory scale are rare because there is limited method to faithfully trigger the rockburst in composite structure. In the light of a novel true-triaxial experiment method, this study applies cyclic dynamic disturbances with large amplitudes to a highly stressed two-layered granite-sandstone composite structure to reproduce the rockburst. Results show that rockburst in layered composite rocks is successfully triggered when subjected to a highly anisotropic true-triaxial loading and cyclic dynamic disturbance. The fragments generated from rockbursts manifest themselves in a spallation failure manner, and the presence of the free surface is vital for inducing rockbursts. This study provides an experimental method for investigations and interpretations of rockbursts in layer composite rocks in deep underground engineering.

Appendix

Two criteria used for determining the preset values of σ₁. Four parallel biaxial compression experiments, with a fixed 30 MPa confining pressure in the X direction, for each MC are conducted here to determine the respective crack damage stress σ_cd. Their complete stress–strain curves are shown in Fig. 7. Stress perturbations such as the stress drop often occur at σ_cd (Zhang and Zhou 2020a), and this characteristic serves as a faithful criterion for identifying σ_cd. Multiple parallel experiments aim at reasonably determining the stress point at which the σ₃-unloading should be implemented. We expect that the stress state at σ_cd in biaxial compression experiments can mimic the stress state at the moment of the σ₃-unloading in true-triaxial tests. This consideration is because (1) the specimen is highly stressed and the stored energy is large, which is believed as an internal factor of rockbursts, and (2) the specimen’s strength is not reached so that the dynamic disturbance with large amplitude can be applied to a highly stressed rock mass to induce rockbursts. The two criteria, including (1) the presence of a large amount of strain energy in rocks and (2) the presence of the specimen’s strength, are utilized for determining the preset values of σ₁ in true-triaxial tests, as listed in Table 2. Figure 7 suggests that the tested materials have a good homogeneity and the experiments have a good repeatability. The expected σ_cd is approximately 58 MPa for MC = 6.48, 126 MPa for MC = 1.47, and 149 MPa for MC = 2.59. Considering the loading–unloading paths in Fig. 2, the preset values of σ₁ in the Z direction should be respectively 43 MPa for MC = 6.48, 111 MPa for MC = 1.47, and 134 MPa for MC = 2.59.

Fig. 7

Complete stress–strain curves of the parallel experiments for each MC