The rock mass deformation modulus (Em) is an essential input parameter in numerical modeling for assessing the rock mass behavior required for the sustainable design of engineering structures. The in situ methods for determining this parameter are costly and time consuming. Their results may not be reliable due to the presence of various natures of joints and following difficult field testing procedures. Therefore, it is imperative to predict the rock mass deformation modulus using alternate methods. In this research, four different predictive models were developed, i.e., one statistical model (Muti Linear Regression (MLR)) and three Artificial Intelligence models (Artificial Neural Network (ANN), Random Forest Regression (RFR), and K-Neighbor Network (KNN)) by employing Rock Mass Rating (RMR₈₉) and Point load index (I₅₀) as appropriate input variables selected through correlation matrix analysis among eight different variables to propose an appropriate model for the prediction of Em. The efficacy of each predictive model was evaluated by using four different performance indicators: performance coefficient R², Mean Absolute Error (MAE), Mean Squared Error (MSE), and Median Absolute Error (MEAE). The results show that the R², MAE, MSE, and MEAE for the ANN model are 0.999, 0.2343, 0.2873, and 0.0814, respectively, which are better than MLR, KNN, and RFR. Therefore, the ANN model is proposed as the most appropriate model for the prediction of Em. The findings of this research will provide a better understanding and foundation for the professionals working in fields during the prediction of various engineering parameters, especially Em for sustainable engineering design in the rock engineering field.

Effective monitoring of rock fracture and seepage is an important information means to ensure the safety of geotechnical engineering. Therefore, sandstone samples were subject to uniaxial compression under different hydraulic conditions in the presence of infrared radiation and observation. This study uses the multiple infrared radiation indexes (ΔAIRT, IRV, VDIIT) and image data to analyze the influence of coupled stress-hydro effect of infrared radiation change on sandstone surface. The main findings are: (1) The surface temperature of sandstone samples rises in the compaction and linear elastic stages, keeps stable or decreases in the fracture development stage, and rapidly decreases in the post-peak failure stage. (2) The samples with internal water pressure not more than 0.30 MPa, surface temperature and load curve at the compaction and linear elastic stage have a strong power function relationship, which a coefficient of determination is 0.8900. (3) The IRV curve appears as a pulse jump at the time of water seepage. After that, both the fracture development and the post-peak failure stages have stepped up. The VDIIT curve also appears to be a pulse jump at the time of water seepage, and obvious up and down fluctuations exist before water seepage and fracture. (4) Based on the Pauta Criterion, by analyzing the values of VDIIT during the experiment, the early warning threshold of sandstone fracture seepage is determined to be 0.00559. The research finding can provide an experimental and theoretical basis for the early warning of flood accidents in underground rock engineering.

The brittleness index is one of the most integral parameters used in assessing rock bursts and catastrophic rock failures resulting from deep underground mining activities. Accurately predicting this parameter is crucial for effectively monitoring rock bursts, which can cause damage to miners and lead to the catastrophic failure of engineering structures. Therefore, developing a new brittleness index capable of effectively predicting rock bursts is essential for the safe and efficient execution of engineering projects. In this research study, a novel mathematical rock brittleness index is developed, utilizing factors such as crack initiation, crack damage, and peak stress for sandstones with varying water contents. Additionally, the brittleness index is compared with previous important brittleness indices (e.g., B₁, B₂, B₃, and B₄) predicted using infrared radiation (IR) characteristics, specifically the variance of infrared radiation temperature (VIRT), along with various artificial intelligent (AI) techniques such as k-nearest neighbor (KNN), extreme gradient boost (XGBoost), and random forest (RF), providing comprehensive insights for predicting rock bursts. The experimental and AI results revealed that: (1) crack initiation, elastic modulus, crack damage, and peak stress decrease with an increase in water content; (2) the brittleness indices such as B₁, B₃, and B₄ show a positive linear exponential correlation, having a coefficient of determination of R² = 0.88, while B₂ shows a negative linear exponential correlation (R² = 0.82) with water content. Furthermore, the proposed brittleness index shows a good linear correlation with B₁, B₃, and B₄, with an R² > 0.85, while it shows a poor negative linear correlation with B₂, with an R² = 0.61; (3) the RF model, developed for predicting the brittleness index, demonstrates superior performance when compared to other models, as indicated by the following performance parameters: R² = 0.999, root mean square error (RMSE) = 0.383, mean square error (MSE) = 0.007, and mean absolute error (MAE) = 0.002. Consequently, RF stands as being recommended for accurate rock brittleness prediction. These research findings offer valuable insights and guidelines for effectively developing a brittleness index to assess the rock burst risks associated with rock engineering projects under water conditions.