Publications

Rock is a typical brittle material, and the evaluation of its brittleness index has important guiding significance for hard rock resource exploitation, unconventional oil and gas resource exploitation, mechanical driving efficiency, rock burst prediction, and dynamic disaster prevention and control. At present, brittleness index often measures the brittleness of rock under static load; thus, whether it is applicable to dynamic load is worth exploring. In this study, static and dynamic uniaxial compression tests and Brazilian splitting tests were carried out on five kinds of rocks, including fine granite, coarse granite, shale, marble, and sandstone, using the INSTRON−1346 test system and split−Hopkinson pressure bar (SHPB), respectively. The brittleness index values of different rocks under static and dynamic load were determined, and the changes in the brittleness of rocks under different loading methods and different strain rates were studied. The definition of brittleness and the applicability of existing brittleness indices were also discussed. It was found that the loading rate amplified the variation of the brittleness characteristics of rock. When static load changes to dynamic load, the brittleness of rocks increases, and the brittleness relationship between different rocks remains unchanged. The more brittle the rock is under static load, the greater the range of brittleness enhancement is under dynamic load. It was also found that the brittleness of sandstone had an obvious effect on the strain rate. The brittleness of rock increases with the increase in strain rate, and the greater the strain rate, the greater the brittleness enhancement degree. These research results can provide reference values for dynamic disaster prevention and safe construction of deep rock projects such as mines and tunnels.

The rockburst prediction becomes more and more challenging due to the development of deep underground projects and constructions. Increasing numbers of intelligent algorithms are used to predict and prevent rockburst. This paper investigated the drawbacks of neural networks in rockburst prediction, and aimed at these shortcomings, Bayesian optimization and the synthetic minority oversampling technique + Tomek Link (SMOTETomek) were applied to efficiently develop the feedforward neural network (FNN) model for rockburst prediction. In this regard, 314 real rockburst cases were collected to establish a database for modeling. The database was divided into a training set (80%) and a test set (20%). The maximum tangential stress, uniaxial compressive strength, tensile strength, stress ratio, brittleness ratio, and elastic strain energy were selected as input parameters. Bayesian optimization was implemented to find the optimal hyperparameters in FNN. To eliminate the effects of imbalanced category, SMOTETomek was adopted to process the training set to obtain a balanced training set. The FNN developed by the balanced training set received 90.48% accuracy in the test set, and the accuracy improved 12.7% compared to the imbalanced training set. For interpreting the FNN model, the permutation importance algorithm was introduced to analyze the relative importance of input variables. The elastic strain energy was the most essential variable, and some measures were proposed to prevent rockburst. To validate the practicability, the FNN developed by the balanced training set was utilized to predict rockburst in Sanshandao Gold Mine, China, and it had outstanding performance (accuracy 100%).

Rock fragmentation is an important evaluation indicator for field blasting operations. This paper applies a deep learning-based method, the Segment Anything Model (SAM), to automatically segment rock fragments. To review the SAM’s segmentation performance, 83 images of rock fragment collected from the mine site were used as the test dataset. Pixel-level accuracy (PA), intersection over union (IOU), and dice coefficient (Dice) were employed to evaluate the model pixel-level segmentation performance. The results showed that the SAM exhibited excellent segmentation performance on the test data (PA = 94.5%, IOU = 94.4%, Dice = 95.4%). The coefficient of determination (R2) values for the 50% and 80% passing sizes (X50 and X80) were 0.970 and 0.991, respectively, which demonstrated that the SAM could achieve high precision measurement of rock fragmentation. Additionally, the effectiveness of the SAM was further evaluated by comparing it to commercial software, and the generalizability of the SAM was verified on two other datasets. The findings revealed that the SAM not only outperformed the Split-Desktop V 4.0 on the test dataset but also achieved comparable accuracy to previous studies on the two other datasets. The SAM could be regarded as a useful tool to provide fast and accurate feedback for field blasting.

The fracture behavior of the disc specimens in the Brazilian test is closely related to the reliability and accuracy of the experimental results. To comprehensively investigate the effect of various loading methods and rock material types on the failure mechanism of the Brazilian discs, five different rock types tested with three typical loading methods were employed in this work. The digital image correlation (DIC) method was applied to record and analyze the strain and displacement field of the specimens during the loading process. Experimental results indicate that the peak load and deformation characteristics of the Brazilian discs are strongly affected by the loading types. The Brazilian test with the Chinese standard is evidently not suitable for measuring the tensile strength of rocks, and the other two testing methods may lead to an invalid failure mode for rock materials with high stiffness and tensile to compressive strength ratio. Furthermore, it revealed that the maximum equivalent stress point of a disc specimen is co-controlled by the material stiffness and its tensile–compression ratio. The present work shows that it is necessary to select a suitable loading configuration for each rock type in the Brazilian test.

This study presents a peridynamic model formulated using the micromodulus function and bond deformation. The model is derived by establishing energy equivalence between a modified virtual internal bond (VIB) and a peridynamic bond. To address surface effects in peridynamics, a stress-based correction method utilizing nodal stress is introduced, enhancing the model’s numerical accuracy. The model was implemented using an in-house Cython code and validated through the following numerical examples: a plate under traction, a plate with a hole under displacement boundary conditions, a uniaxial compression test on granite with a deformation-based mixed-mode bond failure criterion, and a comparison with an existing strain-based peridynamic model. For the plate under traction, the deformation-based method performed similarly to the strain-based model in the loading direction and better in the unloaded direction. The stress concentration obtained from the proposed model (240 MPa) near the hole in the rectangular plate simulation differed from FEM (252 MPa) by 4.7%. The granite test predicted a UCS of 111.88 MPa and a Young’s modulus of 20.67 GPa, with errors of 0.1% and 1.57%, respectively, compared to the experimental data.