Uniaxial compressive strength (UCS) is one of the most important parameters to characterize the rock mass in geotechnical engineering design and construction. In this study, a novel kernel extreme learning machine-grey wolf optimizer (KELM-GWO) model was proposed to predict the UCS of 271 rock samples. Four parameters namely the porosity (P_n, %), Schmidt hardness rebound number (SHR), P-wave velocity (V_p, km/s), and point load strength (PLS, MPa) were considered as the input variables, and the UCS is the output variable. To verify the effectiveness and accuracy of the KELM-GWO model, extreme learning machine (ELM), KELM, deep extreme learning machine (DELM) back-propagation neural network (BPNN), and one empirical model were established and compared with the KELM-GWO model to predict the UCS. The root mean square error (RMSE), determination coefficient (R²), mean absolute error (MAE), prediction accuracy (U₁), prediction quality (U₂), and variance accounted for (VAF) were adopted to evaluate all models in this study. The results demonstrate that the proposed KELM-GWO model was the best model for predicting UCS with the best performance indices. Additionally, the identified most important parameter for predicting UCS is the porosity by using the mean impact value (MIV) technique.

A detailed and fully coupled thermo-hydro-mechanical (THM) model for fractured rock is presented. This model distinguishes itself by harmoniously integrating elastoplastic material behaviour with a continuum damage mechanics framework. Solid matrix displacement, pore and fissure water pressures and temperature of the solid are introduced as the primary nodal variables. The ingenuity of this research is embedded in the intricate coupling of THM processes with plastic deformation and damage mechanics in a double porous medium, a venture that significantly broadens the remit of existing methodologies. The model is implemented using finite element method (FEM) and validation is achieved by comparing the FEM results against existing literature numerical outcomes describing linear and elastoplastic continuum damage behaviour of fractured rock. The model also exhibits an extraordinary proficiency in reproducing experimental triaxial test results, using THM components conjoined with elastoplastic bounding surface aspects and inherent hardening effects. It is imperative, nonetheless, to underscore the model's sensitivity to certain material properties, inclusive of strength parameters, leakage coefficients, and permeability attributes. This fully coupled THM model provides a comprehensive and sophisticated tool for investigating the behaviour of fractured rock under various loading conditions. It can help us better understand the physics of fractured rock behaviour and contribute to the development of more accurate and reliable models for engineering applications, such as CO₂ injection.

An anisotropic plasticity model for the description of nonisotropic mechanical behaviors of porous media is presented. This model is developed using bounding surface plasticity theory within the critical state framework. The inherent and stress-induced anisotropy has been accounted for by an anisotropic variable, which is employed with the consideration of both irrecoverable deformation and stress state. The evolution of anisotropy has been described by both rotational and distortional hardening rules. The rotational hardening rule is proposed for the consideration of bounding surface inclination induced by inherent or induced anisotropy. The distortional hardening rule is developed to capture the varying shape of bounding surface associated with the accumulation of irrecoverable deformation and anisotropy. This model has been validated through modeling the stress–strain responses of porous media under various loading conditions, including triaxial drained/undrained compression/extension tests on both isotropically and anisotropically consolidated samples. Good agreement between model simulation and experimental results has been achieved for all cases taken into consideration, demonstrating the capability of the proposed model.