The hard Flint limestone of Shuibuya hydropower station underground construction cavern is utilized as a research object to investigate the creep problem caused by excavation of rock masses such as caverns. In order to perform a triaxial compression grade-unloading creep test, the actual adjustment path of stress during excavation of underground cavern surrounding rock is used. Limestone under different confistiff pressures is then evaluated. Based on the Nishihara model, the elastic damage element considering time-dependent damage is introduced, and the unloading creep constitutive model of stiff Flint limestone is established and verified by experiments. The results show: 1) Deformation and creep strain appear at all stress levels. 2) As the unloading amount increased from 2 to 4 MPa, the quasi-destructive stresses of the samples were smaller from 83 to 79 MPa, indicating that the unloading amount affected the final creep damage strength of the rock samples. In other words, the higher the unloading amount, the lower the ultimate creep failure strength. 3) When entering the accelerated creep stage, the axial and lateral creep strains of the sample increase non-linearly, and the rupture duration of the sample is very short. Therefore, the creep deformation and creep rate characteristics of this stage should be paid attention to in practical engineering. 4) Different from the loading stress path, the failure mode of the Flint limestone rock sample is different. When the unloading amount is 2 and 4 MPa, the creep failure mode of the Flint limestone rock sample is shear failure, showing a significant oblique section crack. 5) The non-linear creep model curve of aging damage and the fitting effect of the unloading creep test curve are acceptable. The rationality of the established non-linear creep model is illustrated.

Underwater crack repair is challenging due to drainage and exhaust, slurry retention at fixed points, and other issues. Magnetically driven epoxy resin cement slurry was developed, which can perform directional movement and fixed-point retention of slurry under the effect of an applied magnetic field. This paper focuses on slurry fluidity and tensile properties. Firstly, in the preliminary pre-study, the main influencing factors of the ratios were determined. Then, the optimum range of each factor is determined by a single-factor experiment. Furthermore, the response surface method (RSM) is applied to obtain an optimal ratio. Finally, the slurry is characterized by micro. Results showed that the evaluation index F proposed in this paper can well evaluate the interaction between fluidity (X) and tensile strength (Y). The 2FI regression model and the quadratic regression model are developed with fluidity and tensile strength as the response values and Epoxy Resin (ER) content, water-cement ratio, Fe3O4 content and sulphoaluminate cement (SAC) content as the influencing factors, and have reasonable fit and reliability. The relationship between the degree of influence of the influencing factors on the response value X and the response value Y in ascending order was: ER content > water-cement ratio > SAC content > Fe3O4 content. The magnetically driven slurry made by the optimal ratio can reach a fluidity rate of 223.31 mm and a tensile strength of 2.47 MPa. This is with relative errors of 0.36% and 1.65% from the model predicted values. Microscopic analysis showed that the magnetically driven epoxy resin cement slurry had a favorable crystalline phase, surface morphology, and structural composition.