Recent Advances in Rock Mass Engineering

Rock mass engineering serves as a critical foundation for infrastructure construction and resource development. Research and practice in this field are increasingly challenged by complex deep geological environments, heterogeneous geological structures, and heightened demands for safety and sustainability [1]. With urbanization accelerating and the extraction of energy and mineral resources reaching greater depths, rock mass engineering is being applied more extensively in tunnels, mines, underground spaces, slopes, and high dam structures. The scale, depth, and complexity of these projects continue to grow [2]. Traditional empirical design methods, which largely rely on historical engineering experience and simplified assumptions, often fall short in providing sufficient reliability and safety assurance when confronted with complex rock mass structures, multi-directional stress states, groundwater effects, and disturbances such as seismic activity or mining operations [3,4]. Modern rock mass engineering requires designs that not only ensure safety but also consider economic efficiency, construction speed, and environmental sustainability, placing greater demands on rock mechanics theory, experimental techniques, and numerical simulation.

Recent years have witnessed rapid progress in rock mechanics, particularly regarding experimental techniques, numerical simulation, constitutive modeling, and the integration of artificial intelligence (AI). These advances provide new theoretical foundations and technical pathways for predicting rock mass stability, optimizing support designs, and controlling geological hazards. For instance, developments in experimental techniques, such as triaxial compression, rock fatigue testing, dynamic loading, and micro-observation methods, enable precise observation and quantitative analysis of micro-crack evolution, damage mechanisms, and macroscopic mechanical responses in rocks [5]. In numerical simulations, approaches based on the finite and discrete element methods and multi-physics coupling models allow engineers to predict rock mass responses, analyze potential failure modes, and optimize support and construction schemes during the design phase [6]. The advancement of AI technologies offers new tools for data processing, intelligent monitoring, and prediction in rock mass engineering. For example, deep learning can analyze rock microstructures, crack distributions, and failure evolution, enabling rapid and accurate assessment of complex rock mass behaviors [7,8].

This Special Issue focuses on key cutting-edge topics in the field in recent years. It brings together multidimensional research encompassing laboratory testing, theoretical analysis, numerical simulation, intelligent algorithms, and field monitoring, aiming to comprehensively present the latest progress in rock mechanics and engineering for addressing practical challenges. This issue particularly focuses on the mechanisms of rock mass failure, the optimization of support structure design, the development of intelligent monitoring technologies, and the integrated applications of these aspects in tunneling, mining, slope engineering, and underground space development. It is intended to provide a high-level platform for academic and engineering exchange and collaboration, promoting high-quality development in rock mass engineering.

In geological engineering, rock masses are often composed of different rock layers. Interfaces between layers represent natural weak planes that are prone to initiate and propagate interfacial cracks, whose evolution directly affects engineering stability [9]. Research on the failure mechanisms of composite rock masses under dynamic loading not only helps reveal the internal crack propagation laws under mining-induced disturbances but also provides a key scientific basis for optimizing stope layout, designing support systems, and preventing rock bursts and dynamic failures [10]. Studies on the mechanical behavior of rocks containing pre-existing holes further reveal how hole geometry and location govern rock failure modes. Integrated research conducted using uniaxial compression tests, digital image correlation technology, and theoretical analysis shows that stress disturbances induced by holes significantly alter crack initiation locations, propagation paths, and damage evolution characteristics, providing reliable support for the design of underground structures and the evaluation of the stability of surrounding rock [11]. Regarding crack initiation and propagation mechanisms, multi-physical field monitoring combining acoustic emission and CT scanning enables 3D visualization and quantitative characterization of crack nucleation, while theoretical analyses reveal the intrinsic relationship between nucleation behavior and rocks’ ultimate failure modes [12]. In studies of layered rock mass failure, fracture mechanics analyses clearly explain how increasing notch length enhances stress concentration effects, leading to reductions in shear strength and stiffness [13]. This theoretical achievement effectively addresses the difficulty of precisely defining crack propagation mechanisms in layered rock under shear loading, providing a solid theoretical basis for predicting the stability of layered surrounding rock and designing supports in tunnels, mines, and other engineering projects.

With the promotion of green mining concepts, microwave-assisted rock breaking has emerged as an efficient new technology. However, the effects of microwave heating on water-bearing rock masses and their influence on mechanical properties are not yet fully understood. Research in this area is of significant practical importance for the efficient extraction of deep hard rock [14]. The rapid development of numerical simulation technology has facilitated a shift in rock mass modeling from traditional homogeneous rock mass to more realistic gradient rock mass structures, offering new theoretical frameworks for slope stability analysis and the simulation of underground engineering responses [15]. Concurrently, studies on joint network structures indicate there is a significant nonlinear coupling effect between joint strength parameters and connectivity, and this effect plays a dominant role in the evolution of overall rock mass strength [16].

The application of AI in rock mass engineering also shows broad prospects. Researchers have developed a transferable, scalable, and interpretable intelligent fracture recognition framework, providing key technical support for digital rock physics and intelligent geological engineering [17]. Multimodal recognition methods based on deep learning and image-processing technology can automatically identify fractures, mineral distributions, and borehole structures within rock-sample CT images and achieve 3D reconstruction and modeling, offering new tools for the refined characterization of complex rock masses [18]. Furthermore, using deep networks such as ResNet50 for identifying microstructural damage in heat-treated rocks enables automatic discrimination of thermal damage and visualization of the micro-damage evolution process, providing innovative methods for geothermal development and post-fire engineering assessment [19]. Combining discrete element simulation with deep learning, systematic studies have been conducted on the evolution of granular material fabric under different loading conditions and its impact on macroscopic mechanical behavior, demonstrating good physical interpretability and engineering applicability [20].

Rock mass engineering is transitioning from an “empirical–static” paradigm to a “data–dynamic” one. Comprehensive results from experiments, numerical simulations, and field monitoring indicate that only through deep integration of rock micro-damage mechanisms, multi-source monitoring data, and intelligent prediction models can the safe, low-carbon, and efficient construction and operation of underground engineering truly be achieved. This transformation requires researchers to possess interdisciplinary knowledge spanning rock mechanics, structural mechanics, computational science, and data science as well as more flexible design concepts and real-time monitoring methods in engineering practice. We anticipate that these research outcomes will serve as a shared language for the academic and engineering communities, stimulate further interdisciplinary collaboration, and encourage continued exploration of deeper, larger, and more complex challenges in rock mass engineering, thereby revealing new directions for future development.