2.1 Research on the whole process of rock deformation and destruction

Natural rocks contain numerous fissures, joints, fractures, and other structural planes. Due to the existence of these anisotropic weak surfaces, when subjected to the action of external forces, their internal often continue to crack, extend, and penetrate along the weak surfaces. Ultimately, this leads to complete destabilization of the rocks (Liu, 2016; Zhang, 2016). Therefore, the destabilization damage of rocks is more complicated and is part of the problem of complex deformation and destruction (Sun et al., 2011).

2.1.1 Research on the whole process of fractured rock deformation and destruction

In the mining of the ore body and roadway excavation process, the ore body or roadway will produce numerous newborn fissures, new pores, and other defects within the surrounding rocks due to the influence of excavation. The integrity and continuity of the surrounding rocks around the ore body are severely affected by the presence of multiple complex structural surfaces (Fu et al., 2018; Haeri et al., 2014; Yin et al., 2014). Numerous studies have shown that the destructive destabilization of rocks in underground engineering is closely related to their weak structural surfaces such as internal fissures, joints, and cross-sections (Kang et al., 2014, 2016; Lin et al., 2017). As a result, many discussions have occurred on the laws of initiation, extension, and penetration of cracks in fissured rocks. Sagong et al. (2002) obtained the crack extension pattern of a defective specimen by compression testing and found that the crack pattern of a specimen with multiple defects was similar to that of a specimen with two defects. As shown in Figure 3a, two types of cracks are generated from the crack tip: wing cracks and secondary cracks. Wu et al. (2021) studied in depth the anchoring effect of anchors on fractured rock masses and their influence on crack extension. The presence of anchors can effectively limit the emergence and extension of tensile cracks and inhibit the emergence of shear cracks. As shown in Figure 3b, the crack types in the prefabricated fissure without the influence of anchors are diverse and similar to Figure 3a, in that wing cracks and secondary cracks dominate. Wu et al. (2020) analyzed the crack extension paths and fracture characteristics of orthogonal cross-fractured rocks with acoustic emission and surface strain measurement systems. This study also found that the cracks of fractured rock mainly spread in the form of wing or anti-wing extensions. Chen et al. (2022) investigated the impact of prefabricated fissure dimensions on the directional expansion behavior of rock Type-I fissures. They also examined the evolution patterns of acoustic emission and deformation field during crack propagation. Analysis of the acoustic emission trends indicates that enhancing the length of prefabricated fissures enhances the stability of directional crack extension, while widening these fissures leads to instability in the primary crack propagation path.

                 

We can find that all of the above research results are from the macroscopic point of view to explore the evolution law of cracks in fractured rocks because the destruction of fractured rocks is the result of damage evolution with continuous sprouting, extension, and penetration of cracks within them. However, the selection of observation methods is insufficient. Therefore, the digital image correlation (DIC) method based on the mesoscopic angle has been widely used in the research of rock crack extension in recent years. For example, Cen et al. (2014) investigated the extension mechanism of a fractured rock mass under dynamic loading with high strain rates. Based on the parallel bond model (PBM) (Zhang et al., 2012) of mesoscopic particles (Figure 4), three types of displacement patterns between mesoscopic particles and their forms are defined, and the crack extension is generalized into six basic modes. Gao et al. (2015) found that the fracture time, fracture toughness, and crack extension velocity of granite showed loading rate dependence by DIC techniques. Zhang, Xi, et al. (2021) determined mechanical characteristics such as crack initiation, strength, stress–strain, and damage morphology of granite based on the fracture mechanics theory to establish a mesoscopic particle flow model for granite containing prefabricated fissures. Based on the illustration in Figure 5a, it can be observed that the crack inclination angle is 30°. The deformation damage process in granite can be classified into four distinct stages, namely, crack closure, elasticity, crack stable extension, and crack unstable extension. Furthermore, Figure 5b displays the axial stress values that correspond to the crack's volumetric strain. The total volumetric strain and crack volumetric strain curves are utilized in this case to identify the crack initiation stress values of the rock samples. Wang et al. (2023) analyzed the relationship between abrupt changes in shear strain field and crack extension within open rocks by using precursory indices of full-field strain information in DIC technology. The results show that the changing patterns of strain field absolute standard deviation and the absolute variation coefficient were highly correlated with the rock fracture evolution. All the above experimental and research studies fully reflect the whole process of internal damage evolution of specimens using the DIC technique observation. However, the above studies lack a quantitative description of the evolution of the crack displacement field during rock fracture by the DIC technique. It can intuitively determine the fracture types of rocks with filled and unfilled fissures, overcoming the limitations of the traditional experience-based judgment of rock fracture types (Miao et al., 2021; Zhang et al., 2023). Future research on the deformation damage of defective rocks should combine macroscopic and mesoscopic aspects, together revealing the whole process of rock deformation damage.

                 

                 

2.1.2 Research on the whole process of rock with filled fissure deformation and destruction

In actual underground engineering, there will be fissures, joints, fractures, and other structural surfaces within the rocks. When the rock mass is disturbed, its weak surfaces are susceptible to deformation and fracture, and weak filling material can also occur within the joints. As a result, the bearing capacity and deformation characteristics of the filled fractured rock mass are significantly different from those of the unfilled fractured rock mass (Bai et al., 2020; Chang et al., 2022). As the filling gradually infiltrates into the surrounding fractured rock masses, it induces a significantly lower state of stress concentration near its weak face (Li, Wang, et al., 2011). Accordingly, fissure filling has become a crucial method to improve the physical–mechanical properties of the fissure weak surfaces and improve the integrity and carrying capacity of the broken surrounding rock in underground filling mining technology to be able to inhibit the rupture of the rock mass to some extent. Yin et al. (2016) conducted a study to examine the impact of different types of filling materials on the mechanical properties and crack growth evolution of fractured rocks. The researchers found that low-strength filling materials (Types 1–3) led to the development of tensile cracks from the tip of the fissure, which connected the left tip through the main fissure and the outer tip of the secondary fissure. Ultimately, this resulted in destabilization and damage to the sample. However, high-strength filling materials (Types 4 and 5) caused secondary cracks to sprout from the two tips of the primary cracks and extend through the upper and lower end surfaces of the sample, leading to destabilizing damage (Figure 6a). Based on the correlation method of digital images, Si et al. (2019) explored the effect of the consolidation time of the fillers on the damage process of fractured rocks from a mesoscopic perspective. They found that the crack forms dominating the damage were inconsistent under different consolidation times. Li, Du, et al. (2021) investigated the effects of fracture, filling, ligament length, and bridge angle on rock strength, fracture pattern, and permeability under hydraulically coupled conditions. They found that the influences of filling material on crack extension and coalescence behavior increased with decreasing rock bridge angle and increasing ligament length. Furthermore, sand filling usually has a greater impact on rock strength properties than mud filling (Figure 6b). Luo et al. (2021) conducted compression–shear experiments on rock specimens containing three different types of cracks. These cracks were either unfilled or filled with various materials. They used the discrete element method (DEM) to analyze the mechanical properties and damage patterns of the cracked rock specimens. The findings demonstrated that the mechanical properties of the specimens and the extent of crack propagation were primarily influenced by both the strength of the fillers and the inclination angle of the fissures. In a related investigation, Wu et al. (2023) focused on the phenomenon of tensile damage in naturally filled rock masses. They used 3D printing techniques to create rock materials with artificially filled fractures. They discovered that different filling conditions led to shifts in the location of the high-strain zone within the rock by digital imaging techniques. In addition, the displacement field of the filled rock transitioned from being continuous to discontinuous.

                 

From the above research, we found that the mechanical properties of rock with filled fissures mainly hinge on the intensity of the filling materials and the inclination angle of the fissures. In terms of damage patterns, the mechanism of the fillers involves weakening the stresses on the defective rock, and the greater the reduction of shear stress, the more easily the filled rock samples will have tensile–shear damage. In contrast, when the strain is substantially reduced, the damage pattern of the filling rock samples is comparable to that of the intact rock samples.

The fillers trapped in the fissures cause the fractured rock to form a massive weak plane structure. Under the influence of stress paths, the filled fractured rocks are subjected to compaction, strength changes in the vicinity of the fissures, and the formation of precracks in the tips of the cracks. In addition, the presence of the fillers also results in a slight increase in the compressive strength of the rocks. Yuan et al. (2018) found that compared with rock samples without filling, rock samples with filling have higher peak strength. As the filling materials change, the peak intensity also changes (Figure 7). Under the influence of the filling bodies, the rock with filled fissures is dominated by wing cracks as a form of damage at the macroscopic level and shear damage at the mesoscopic level. Thus, under uniaxial compression conditions, the rock samples will first undergo a brief compression-dense phase before entering a linear-elastic phase, regardless of whether it is filled or unfilled. When the peak strength is reached, the pressure falls quickly, at which time the rock sample is seriously damaged, and most of its destruction is due to the combined effect of shearing and tensioning. Therefore, the crack generation law for them is that initially there are no micro-cracks. Subsequently, micro-cracks start to emerge at the contact part between the filled bodies and the joint surface as well as inside. Then, the amount of micro-cracks grew rapidly and gradually evolved to the internal rock samples. Finally, micro-cracks gradually develop through the formation of macro-cracks, resulting in destabilization of the rock samples' damage (Cui et al., 2023; Wang, Fu, Song, et al., 2020; Yin, 2023).

                 

In summary, the current research on the deformation damage of fractured rocks has yielded a certain system. However, research on the deformation damage of rock with filled fissures is still in the development stage. Regardless of whether the fractured rock is filled or unfilled, most of the existing research is carried out in terms of the stress–strain curves obtained from the mechanical tests and the pattern of crack presentation under different fissure modes (Fan et al., 2018; Morgan et al., 2017; Yang et al., 2017). Nonetheless, with the application of acoustic emission, DIC methods, CT scanning, and other new auxiliary equipment, experimental research and analysis methods have improved. It is possible to study the whole process of rock deformation and destruction in detail from a microscopic perspective. We can use new theoretical methods and test technology to investigate the mesoscopic physical mechanism of rock deformation and destruction at a deeper level. As a result, it reflects the macro-mechanical properties of rocks through the mesoscopic perspective and constructs the relationship between macroscopic and mesoscopic aspects. This lays the foundation for continuous development and improvement of the mechanical mechanism of rock joints.

2.2 Mechanical properties of rocks with filled joints

In rock engineering, jointed rock is a prevalent and intricate medium. It has characteristics such as heterogeneity, anisotropy, and discontinuity. The deformation, strength, damage, and rheological properties of such rock significantly influence various aspects of rock engineering, including design, construction, operation, stabilization, and reinforcement strategies (Xu et al., 2005). However, in natural settings, most natural joints are filled joints. This phenomenon arises due to prolonged exposure of rock joints to weathering, erosion, and other physical effects, coupled with misalignment between joint walls. These factors lead to the accumulation of specific thicknesses of filling materials within the joints, thus giving rise to what are known as filled joints. As a result, the type and thickness of the filling medium emerge as critical factors that exert a significant influence on the mechanical characteristics of naturally filled jointed rocks (Sun, 1988; Yang, Wang, et al., 2016). This section summarizes the results of previous experiments on the mechanical properties of rock with filled defects, as shown in Table 1. The study focused on brittle materials, such as sandstone, granite, marble, and rock-like materials (gypsum, concrete, and resin), shaped as mainly cubes and cylinders. Most precast filled defects were fissures, holes, and combinations of defects. The filling materials were gypsum, cement, and resin. The number of defects in most of the studies was 0–3. The defect size was 0–30 mm, the width was 0.8–4.0 mm, and the inclination angle was 0°–90°.

Current studies on the mechanical properties of natural filled rock joints are generally conducted through laboratory experiments, theoretical analyses, and numerical simulations to explore the effects of different filling medium types, filling thicknesses, and prefabricated joint geometries on the strength of filled rock joints (Liu et al., 2014; Wang, Fu, Yang, 2020; Xu et al., 2018; Yu et al., 2023). First, the type of filling medium directly affects the mechanical properties of nodular rocks, and different types of filling mediums trigger different mechanical responses. Common types of filling medium, besides those shown in Table 1, include angular gravel and alteration minerals containing water, among others (Cotecchia et al., 2015; Tian et al., 2017; Wang, Zhang, et al., 2020). Second, the thickness of the filling medium shows a significant correlation with the contacting state of the upper and lower rock walls adjoining the joints. Consequently, this thickness variation influences the mechanical attributes of these joints. On the one hand, when the filling medium is thin, the interaction between the upper and lower joint walls during the stress process is affected by not just the wall-filling medium interaction but also the joint surface morphology. Under these circumstances, a connection exists between the strength of naturally filled joints, the thickness of the filling medium, and the morphology of the joint surface. Hence, the conventional theory about the strength of unaltered rock or soil joints is not applicable in this context. On the other hand, the direct contact between the upper and lower joint walls is missing when the filling medium is thin. At this time, the mechanical attributes of the joint are chiefly governed by the inherent properties of the filling medium. This influence is independent of the surface morphology and the strength of the joint walls. Consequently, we can draw upon pertinent theories from the domain of soil mechanics for research and discourse in this scenario (Chai et al., 2020; Li et al., 2022).

Under uniaxial compression conditions, the full stress–strain (σ–ε) curve can effectively reveal the mechanical and deformation characteristics of the rock with filled joints. As shown in Figure 8a (εd, εv, and ε1 are the radial, volumetric, and axial strains of the rock, respectively), the deformation of the rock with filled joints can be roughly categorized into five stages: (I) fracture pore compaction stage; (II) elastic deformation stage; (III) crack stable development stage; (IV) crack nonstable development stage; and (V) postrupture stage. In the early stage of loading, the original tensile structures or microfractures inside the rock gradually close and the rock is compacted, forming early nonlinear deformation. Also, the σ–ε curve is concave upward. In the elastic deformation stage, the σ–ε curve is approximately linear. In the crack stable development stage, the slope of the σ–ε curve tends to decrease with the increase in stress, and new microcracks are generated inside the rock and develop stably within the rock. With the increase in load, the σ–ε curve becomes convex upward, indicating that the cracks inside the specimen develop unstably until the specimen completely ruptures. In the postrupture stage, the internal structure of the specimen is damaged, but the specimen remains essentially intact. As shown in Figure 8b, the filling material and the fissure inclination angle significantly influence the mechanical properties of the specimens. The presence of filling material significantly increases the strength of the defective specimen compared with the unfilled material.

               

The deformation of rocks with filled joints depends not only on the intrinsic properties of the rock but also on the development of microcracks during rock deformation. The deformation and destruction of rocks with filled joints are accompanied by crack closure, sprouting, expansion, and penetration. The pre-peak deformation phase of the rock includes four important characteristic stress thresholds, that is, crack closure stress (σc), crack initiation stress (σi), crack damage stress (σd), and peak stress (σp), as shown in Figure 9. The volumetric strain and microcrack volumetric strain curves in Figure 9 show that the specimens undergo compression, followed by expansion. This is mainly because the microcracks inside the specimens are compressed and dense in the initial stage of loading. Further, after the elastic deformation stage, the cracks sprout, expand, and interact with each other, causing the rocks to expand and enlarge overall.

               

In summary, the main characteristics of filled jointed rock samples are presented in Table 1, and Figures 8 and 9. Most of the prefabricated fissures studied so far are smooth, single fissures. However, the fractures in natural rock masses are mostly irregular, anisotropic, and discontinuous. Research on prefabricated fissures with rough surfaces needs improvement. Also, the effect of filling materials on the rocks with filled joints should be further explored. Previous data show that many types of filling materials exist, and the mechanical properties of the specimens are significantly affected by whether the fissures are filled or unfilled materials. In addition, the impact of the environment around the rock mass is not considered. This mass, in some cases, is subjected to acidic and alkaline conditions over a long period, which drastically affects the physical properties of both the filling material and the rock mass itself.

2.3 The influence of shear on filled rock joints

For jointed rock masses, shear is also an important factor in their stability. It is mainly affected by its size, roughness, strength, the presence or absence of fillers, and other properties and working conditions. Hence, experts both here and abroad have dedicated considerable efforts to investigating the shear mechanical attributes of unfilled joints. The findings from these investigations provide comprehensive insights into their mechanical characteristics and the progression of joint damage. Empirical equations for deformation and shear strength were finally obtained. Among them, Barton (1973) proposed the renowned joint roughness coefficient-–joint compressive strength (JRC-JCS) equation, established on the basis of an extensive series of direct shear tests. While this equation finds widespread implementation within engineering applications, its utility is somewhat curtailed due to the inherent subjectivity characterizing the parameters embedded within the equation itself. Esaki et al. (1999) and Olsson et al. (2001) verified and improved the above empirical Barton's equation utilizing coupled shear–seepage tests of rock joints in the laboratory. Subsequently, through experimental investigations, Grasselli (2006) introduced the initial peak shear strength equation, uniquely tied to the three-dimensional characteristics of the nodal surface morphology. This study advanced the research of nodal shear mechanical properties forward to a deeper level. Soon after, Tang et al. (2012) modified the equation based on several straight shear test results of Grasselli (2006) and its shape parameters. The new equation conforms to the Mohr–Coulomb friction law and the fitting results have smaller errors from the test values. Currently, research efforts in shear test studies about unaltered joints are centered on strength equations reliant upon three-dimensional morphology parameters, along with adjustments to the mechanical parameters encompassed by these equations (Tang et al., 2014; Xia et al., 2014; Yang, Rong, et al., 2016). Building upon the foundational equations, numerous researchers have devised shear strength models for unfilled coupled joints. These models serve to unveil the shear strength characteristics of jointed rocks more comprehensively (Shan et al., 2021; Tian et al., 2018).

Natural rock joints are usually not closed, and a definite thickness of the filling material will generate filled joints. The properties of filling materials play an essential role in controlling the joints' strength (Jiao et al., 2018, 2019; Xu, Lei, et al., 2019). Therefore, some researchers have investigated the peak shear strength of rocks with weak filler joints (Davies et al., 2000; Indraratna et al., 2005, 2008). In their study, Sun et al. (2014) investigated the peaking shear intensity of rock joints filled with cement paste using theoretical analyses and in-house straight shear experiments. They managed to derive the pattern of peaking shear intensity in these joints under varying filling degrees (Figure 10). The filling model can describe the change rule of peak shear strength of fractured rock joints after filling cement slurry. At the same time, it can reflect the different mechanical effects of filling cement slurry on hard rock and soft rock joints. Wu et al. (2019) conducted a straight-shear mechanical test study on the specimens of the filler–rock contact surface. Linear correlation conversion relationships were obtained for the friction angle and cohesion within the contact surface with the friction angle and cohesion within the isotropic ratio and age fillers, respectively. The established straight shear test setup is shown in Figure 11: Figure 11a shows the schematic diagram of the straight shear instrument. Granite is used as the rigid filler in this experiment, and the straight shear mode is shown in Figure 11b. Zhong et al. (2020) took into account the heterogeneity of filling materials and generated rock samples of filling mortar joints with initial fractures of different lengths (Figure 11c). The compression–shear fracture mechanism and initial fracture size of rock mass with fissure-filled joints under uniaxial compression are studied, and the influence of fracture mode and fracture energy on rock mass is revealed.

               

               

From the above research, we can find avenues for future research on the shear intensity of filled rock joints: consideration of natural rock joints as the research subject, in-depth study of the theory of shear strength micro-mechanics, preparation of specimens with filled joints, shear effect, etc. Modeling shear strength is more consistent with naturally filled rock joints. Correspondingly, Xiao et al. (2020) prepared three sets of thin-layer filling rock joint specimens with Barton's standard profile line morphology characteristics and five different filling degrees. They investigated the shear expansion characteristics of thin-layer filled rock joints and found that the surface roughness of the joints is the key factor affecting their shear expansion characteristics. An analysis of the effects affecting the shear expansion characteristics of thin-layer filled rock joints is established, as shown in Figure 12. In a study conducted by Jiao et al. (2022), the impact of thickness-filled with sand on the intensity of shear in joints was examined. The researchers discovered that the filling significantly influences the mechanical properties of shear by regulating the damage mode of the joints. They also described the obtained trends in the cohesion and internal friction angle of the filled joints, as shown in Figure 13. Analysis of the data reveals a slight increase in the internal friction angle of the filled joints compared to the unfilled ones. Conversely, the cohesion of the filled joints shows a noticeable reduction, with the overall decrease in cohesion being higher than that of the internal friction angle. Another investigation by Wu, Zheng, et al. (2022) used PFC2D to examine how joint roughness and filler thickness affect the shear characteristics of joints. The numerical outcomes indicate that an increase in joint roughness and a decrease in filler thickness result in elevated shear stress and normal displacement in the joints.

               

               

In summary, we found that most of the rock specimens in the aforementioned study were cut to obtain prefabricated fissures with smooth surface structures and a certain interface with the fillers. However, the surface structure of actual engineering rock fissures is highly complex. We should optimize the theoretical model and conduct new experiments with advanced observation instruments. Further, we should optimize the physical model to establish a rock model that is more consistent with natural rock joints. Ways to strengthen the adhesion between the fillers and the prefabricated fissure still need to be further explored when investigating the damage characteristics of rock with filled fissures. Additionally, the effects of multiple factors (e.g., fissure opening, length, roughness, surface area, etc.) on the mechanical properties of rock with filled fissures should be comprehensively considered based on theoretical knowledge and advanced equipment. Finally, we must combine macroscopic and microscopic analyses to fully understand the fracture damage mechanisms of defective rocks.

3.1 Theoretical analysis of seepage on the rock with filled fissures

In natural fractures, filled fractures tend to be prevalent, while unfilled fractures are rare. The permeability behavior of fractured rocks with and without fillers has significant variability. Consequently, study of seepage characteristics of rock with filled fissures is vital and relevant for rock engineering. Currently, there are few research results on the seepage characteristics of filled fractures in terms of research here and globally. However, research on the seepage characteristics of pore media can be considered to be quite adequate, and a certain theoretical system has been developed.

Based on the theoretical system of seepage characteristics of pore media to carry out seepage tests on the filling fissure, a few researchers have obtained a series of seepage equations (Table 2). Kozeny J. has proposed the capillary seepage equation, which shows that the resistance of a liquid passing through filling sand particles is inversely proportional to n3/(1 – n)2 in the case of low-velocity flow. In 1927, Carman P. C. proposed the motion equation of liquid flowing through packed sand in a circular pipe at low speed, thus establishing the more classical theory of the Carman–Kozeny capillary seepage model. This theory laid the foundation for the study of seepage in filled fissures (Vaclav et al., 1979). In 1957, Mintz derived the equation for the permeability coefficient of pore media based on the Carman–Kozeny model (МИНЦ et al., 1957). In 1989, Tian et al. (1989) studied the relationship between the fissure seepage pattern and the filling medium by using the theory of equal hydraulic radii to transform the equation of motion of Carman P. C. for filling sand into an equation for filled fissure seepage. In 1994, Su et al. (1994) established a schematic diagram for seepage analysis of a filled fissure, as shown in Figure 14a: consider the width of the fissure as b, the height of intercepted fissure as t, and the length as l. It is assumed that the porosity of the fillers is m, and the filling material is a spherical body with uniform particles and particle size d. The average flow velocity of the filled fissure is derived according to the capillary seepage model, and the relationship between the permeability coefficient Kf of the filled fissure and the width of the fissure, the particle composition of the filling material, and the porosity of the filling is obtained. From the equation for calculating the permeability coefficient Kf of the filled fissure obtained by Su et al. (1994), it is evident that the permeability of the pore media is independent of the size of the specimen. It depends only on particle composition and porosity. Hence, for nonuniform grain size filling materials, we can incorporate the values of grain sizes d10–d20 into this equation for calculation. This indicates that small particle size particles mainly control the permeability of the pore media with a certain porosity.

               

In Table 2, Vaclav et al. (1979) proposed that the motion equation of the filling sand is the seepage model for pore media. However, there is some similarity between this equation and fissure media in terms of structure. This also suggests that the early studies of seepage characteristics of pore media have laid an important foundation for the exploration of seepage in filled fissures. According to the seepage equations proposed by МИНЦ et al. (1957), Tian et al. (1989), and Su et al. (1994), it is evident that the permeability of the filled fissure is not only related to the porosity of the fillers and the width of the fissure but also there is an obvious correlation with the particle size distribution of the fillers. Some researchers in China have also explored this and proposed a modified relational equation for the cubic law of fissure seepage (Xu et al., 2022). It is believed that the cubic law can also be used to describe the filled fissure seepage, and its correction coefficient is related to the porosity of the fillers. From the modified form, the filled fissure seepage is affected by both the porosity of the filling medium and the width of the fissure. However, the particle size of the fillers does not appear in the correction factor of the cubic law, which is also an avenue for future research.

In summary, we found that the seepage characteristics of rock with filled fissures are closely related to the fissure width, roughness, and physical properties of the filling medium (e.g., particle composition, porosity, and particle size of the filling material) based on previous theoretical analyses. However, there are various types of filling materials in the fissures of natural rock mass, and different properties of filling materials have various impacts on the seepage characteristics of the fractured rock mass. Nonetheless, there are fewer research results on the nature of fissure filling medium. In the study on seepage characteristics of rock with filled fissures, we should attempt to use the natural filling medium. At the same time, we also should strengthen the research on the connection between the physical characteristics of the filling medium and seepage.

3.2 Experimental research

With the progress of science and technology in recent years, the problem of seepage in filled fractured rock mass has been studied in various fields. Accompanied by the application of 3D printing, CT scanning, and laser scanning technology in rock engineering, the distribution, morphological features, and topological relationships of rock fissures can be accurately described, which provides technical support for constructing rough fissure models (Fereshtenejad et al., 2016; Jiang et al., 2016; Larmagnat et al., 2019; Wang et al., 2018). Numerical analysis methods, such as the finite difference method (FDM), the finite element method (FEM), the lattice boltzmann method (LBM), molecular dynamics method (MDM), and smooth particle hydrodynamics (SPH), have been applied to the study of permeability characteristics of fractured rock mass, and a lot of advanced research results have been achieved (Dong, 2020; Li, 2019; Liu, 2021; Yang et al., 2015; Yin, 2017).

Some researchers have conducted experimental studies on filling fissure seepage and its characterization. Liu et al. (2010) investigated the seepage characteristics of sandstone fissures containing dense original rock fillers. They found that confining pressure has a significant effect on the seepage flow of rock samples with filled fissures, but has minimal effect on unfilled fissures with large openings. They also found that the nonlinear change in seepage flow was more pronounced with increasing fracture depth. Liu et al. (2019) analyzed the permeability and strength characteristics of filled fissure sandstones by indoor triaxial seepage tests. They found that the peak permeability coefficient of the filled fractured rock appeared before the peak stress. Chen, Zheng, Zhang (2019) conducted experiments on the fissure permeability characteristics of pulverized coal-filled coal rock with different particle sizes of pulverized coal as the filling medium and with different air pressures and fissure widths under certain confining pressure conditions. They found that the filling width greatly influences the permeability of coal rock fissures with the increase of air pressure under certain confining pressure conditions. Tan et al. (2020) focused on analyzing the influences of two factors, fissure roughness and gap width, on the seepage characteristics of through-filled fissures under confining pressure conditions. Figure 15a demonstrates the standard deviation of permeability for three gap widths of 10 levels of roughness through-filled fissures under the same level of perimeter pressure. It is clear that the effect of the perimeter pressure on the permeability of the through-filled fissures is predominant relative to the fissure roughness and gap width. It can also be seen from the fitted curves in Figure 16a that the permeability tends to decrease nonlinearly with increasing perimeter pressure at different roughness and gap widths. Zhang, Chen, et al. (2021) investigated the seepage characteristics of incompletely filled fissures with brittle-plastic filling materials and different gap widths. They compared the stress–seepage curves of fractured rocks containing brittle filling materials with those containing elastoplastic filling materials and found that the two curves were similar overall. Wang, Zhang, et al. (2022) prepared rock specimens with different non-equal width ratios of filled fissures indoors to study their seepage characteristics under high and low confining pressures. They found that the permeability was quadratically related to the non-equal width ratio under the confining pressure-seepage conditions, and the seepage pattern affected by the non-equal width ratio showed nonlinear characteristics. Ye et al. (2022) studied the physical phenomena of fissure seepage and nonlinear diffusion in mudstone blocks based on X-ray visualization experiments. As shown in Figure 16b, they obtained the mudstone hydro-damage coefficient and the seepage diffusion coefficient by Boltzmann function fitting, and it is evident that the hydrodynamic action exacerbated the damage degree of mudstone, and the seepage diffusion coefficient showed a nonlinear decreasing trend. Gan et al. (2023) investigated the effects of different roughness and filled gap widths on the seepage characteristics of a rough single fissure in limestone based on constant confining and seepage pressure conditions. As shown in Figure 15b, the evolution of seepage volume of different JRC fissure specimens can be divided into three periods: development, transition, and stabilization. They found that the equivalent hydraulic gap widths of the fissures with different filling gap widths decreased with the increase of infiltration time combined with 3D scanning and engraving techniques. From the above studies, the current experimental research on seepage in rock with filled fissures is more focused on the seepage characteristics under the action of the surrounding pressure. In addition to the influence of the surrounding pressure, the fissure roughness, gap width, and types of filling material are important factors that affect the seepage in the filled fissures. However, when the fissure roughness and gap width are small enough, the influence of these factors is slowly decreasing compared with the sufficiently high surrounding pressure.

               

               

In summary, the positive and negative effects of the variability in the filling medium on fissure permeability have been less explored both here and globally. Hence, the structure of the filling medium, fissure characteristics, seepage characteristics, hydraulic coupling, and other factors should be intensively investigated. In addition, NMR can be used to determine the permeability, porosity, and other fundamental physical characteristics of the rocks and fillers. During the seepage tests of rocks with filled fissures, NMR can aid in quickly, accurately, and intuitively observing the seepage process and change of fluid flow inside the defective rock. This approach provides new insights for in-depth exploration of the microscale to mesoscale pore structure of defective rocks.

3.3 Numerical research

With the rapid development of digital science, there have been new methods and ideas for the in-depth study of seepage on filled fissure rocks over the past few years. Researchers, both domestically and globally, have carried out many investigations on the theory of fissure seepage mathematical modeling and numerical analysis methods, and have achieved fruitful results. However, there are relatively few numerical studies on seepage in filled fissures. This subsection briefly summarizes the seepage models and numerical analyses of filled fissures used by domestic and international researchers.

From the experimental studies mentioned above, we can find that the visualization of laboratory tests is weak, the sealing cost is high, and it is difficult to carry out repeatability tests. Therefore, considerable research has been carried out on filled fissure seepage modeling as well as numerical analysis. Zhang et al. (2008) used the three-dimensional FEM to investigate the influence of the hydraulic effect of filling on fracture deformation. In the natural state, the interconnection of horizontal and vertical fissures is the main reason for the large-scale collapse of the filled fissure zone. However, this analysis only considers the weakening effect of the filling medium when exposed to water and does not take into account the swelling effect of the fillers. Since different fillers have unique properties, the conclusions drawn may differ from the actual situation. Therefore, Chen et al. (2012) conducted numerical simulations on the full-filled rock mass of the fissure using finite element analysis software, and the results showed that the expansion effect of the fissure fillers increased the components of each stress in the fractured rock mass. Zhao et al. (2018) utilized Fluent software to analyze the flow behavior of water within three distinct pore structures and unveiled the mechanism behind the transformation of seepage within the fillers occupying the pore spaces in fractured rock formations. To simulate a fault-broken rock mass, a filler architecture was constructed, comprising a fixed granular phase, an active granular phase, and a water phase. Different densities of these structures were established, as depicted in Figure 18a–c. The findings indicated that the impact of gravity on water-phase seepage intensifies as the filling structure becomes looser. Conversely, as the filling structure becomes denser, the diffusion range of the water phase broadens while attaining a more uniform distribution throughout the system. Shao et al. (2020) examined the influence of JRC and uniformity of the filling medium on seepage evolution. To create an artificial fissure, a wavy design was used (Figure 17a), and the corresponding boundary conditions are presented in Figure 17b. The study revealed that the seepage characteristics of the filled fissure undergo a nonlinear transformation, characterized by three distinct stages: slow change, rapid change, and stabilization. The reliability and feasibility of the numerical model are verified by the physical model in Figure 17c,d and the experimental-simulation results in Figure 17e. Yin et al. (2021) investigated the evolution characteristics of seepage over time in a large rock mass containing filled joints using hydrodynamic experiments and the finite element software COMSOL. The above research effectively predicted the characteristics of pore water pressure and flow rate in a fracture-filled rock mass, providing guidance on preventing water surges. In addition, the models mentioned above are subject to further in-depth study for the selection of parameters and the establishment of an isomorphic model.

               

Some researchers have made improvements to the existing model in order to investigate the seepage characteristics of rock containing filled fractures. In a recent study, Liu et al. (2021) developed a novel statistical model of damage based on the equivalent modulus of elasticity of fractured rock. They used this damage statistical model to describe the seepage damage process and damage characteristics of the filled fractured rock masses. They found that the model accurately predicted the influence of different stress conditions and filling rupture states on the strength properties of the rock, which was consistent with experimental data. In another study by Sun et al. (2022), three-dimensional models of fault rock samples were created using CT scanning. The researchers established a seepage model for porous media with various natural fracture structures. Through simulations using Fluent software, it was observed that the pressure gradually decreased along the seepage direction, while the seepage velocity followed a pattern of initially increasing and then decreasing. To investigate the formation mechanism of seepage-damaged water-surge channels and particle transport in sandstone fault fillers, Cai et al. (2022) utilized a coupled EDEM–Fluent computational method. The model that they developed is illustrated in Figure 18d. The seepage process was divided into three stages: slow seepage, sudden seepage, and stable seepage, under variable hydraulic gradient conditions. The advantage of this model is its ability to predict the overall movement trend of particles and observe the relationship between interparticle contact amount and the mass of lost particles. The above models for the study on rocks containing filled fissures are not sufficiently comprehensive and do not allow for the establishment of rock masses consistent with natural environments. In the natural environment, rocks are defective, with rough, non-smooth surfaces. Therefore, we should consider models that are consistent with natural rock masses.

               

In summary, we found that there have been relatively few numerical modeling analyses of seepage in filled fissure rocks compared to seepage in rocks with fractures. Factors such as the complexity of their defects and different pore microscopic structures increase the difficulty of creating this model. All of the above research methods have obtained relatively close to actual results in the analysis of actual fractured rock engineering, but these methods still have some limitations. For different engineering fields, these methods are not widely applicable. However, we can achieve the goal of solving problems in different engineering fields by combining the advantages of these methods. Therefore, integrating and complementing multiple numerical analysis methods with each other represent avenues for future research.

4.1 Effect of stress on the permeability of filled fissures

The analysis of stress–seepage behavior within fractured rock masses is a prominent focus across various fields of engineering and has remained a significant topic in exploration of rock mechanics over the past two decades. Investigation into alterations in stress, strain, and permeability within a dynamically balanced system also holds considerable sway over project safety and stability (Guo et al., 2019). As a subsurface effect, confining stresses bring about a reduction in pore size within fissures, subsequently influencing the porosity of the rock and its associated filling materials. However, the intricate interplay of multiple factors, including the morphology of the fissures, their orientation relative to the principal stress direction, the mechanical and chemical attributes of the rock matrix, and the influence of effective and shear stresses, complicates the overall understanding. These factors together determine the pore size distribution of fractures and the degree of reduction in the porosity of the filler. This leads to an increase in the degree of tortuosity of the seepage channels inside the filling body, which reduces the permeability coefficient of the filling body and has an effect on seepage (Wei et al., 2021; Kacimov et al., 2019; Wu, Wang, et al., 2022; Yin et al., 2020).

The coupling between the fracture wall and the fluid flow may arise along the direction of the filled fissure with the stress field acting on the rock. This coupling may alter the potential mechanical properties of the filled fracture. The opening and closing caused by normal stress and the expansion caused by shear may cause changes in the fracture aperture. According to the research, we found that when excessive loads exceeding the strength of the matrix with the filling body occur, sufficient spaces and channels appear in the rock matrix itself, in the filling body, and in the place where the filling body is in contact with the surface of the fissure. As a result, these expanded permeability channels increase the fluid permeability. In the following, the stress–seepage coupling effect of the fissure fillers is discussed from two aspects, namely, fully filled and partially filled.

4.1.1 Stress–seepage coupling on fully filled fissures

The types of fissure filling mediums are complex and diverse, and different filling mediums have different properties in nature. When subjected to principal stresses, the seepage characteristics between the fracture-filling medium and the filling medium itself are different. The study of stress–seepage coupling was proposed as early as the 1950s (Wu et al.,     1995). Louis first investigated stress–seepage coupling in nondestructive rock masses under their natural state by the 1970s, and his suggestion of a negative exponential relationship between positive stress and permeability coefficients was confirmed by other researchers (White,     1988). In the 1980s, researchers Noorishad from the United States and Ohnishi from Japan raised the issue of coupling of stress and seepage fields (Kelsall et al.,     1984; Liu et al.,     2000). At the beginning of the 21st century, Chinese researchers Jianxiu Wang and Liping Li proposed modeling studies based on stress–seepage coupling in different research contexts (Li, Li, et al.,     2011; Wang et al.,     2008). Subsequently, other researchers have proposed characterization of the behavior of the filling body under the influence of multi-field coupling (Ghirian et al.,     2013,     2014). The existing filled fissure test device only considers the influence of the composition with the filling medium itself on the seepage characteristics, the sealing effect of the test system is poor, the data analysis of the monitoring system is not timely, and so on. On the other hand, some researchers proposed the development of a triaxial permeability test system for research and analysis of stress–permeability coupling and its application in permeability characterization tests of filling medium (Li et al.,     2017). Later, Li et al. (     2020) established a multi-field coupled temperature–permeability–mechanics–chemistry model framework for tailings based on the classical Biot theory of pore elasticity. Figure     19 displays the chemical–physical processes analyzed in the model, elucidating the impact of the multi-field coupling effect on the excessive pore water pressure of tailings. Therefore, the rate of change of porosity under one-dimensional conditions can be obtained as

      $$\frac{\partial n}{\partial t}=(\alpha -n)\left\{\left[\frac{(\alpha -1)}{E_0(t_e)}{\gamma }_w-\frac{1}{E_0(t_e)}{\gamma }^{\prime }+\frac{1}{K_s}{\gamma }_w\right]m+\left[\frac{\alpha }{E_0(t_e)}+\frac{1}{K_s}\right]\frac{\partial u}{\partial t}+\left[\frac{K_d(t_e)}{E_0(t_e)}-1\right]{\beta }_s\frac{\partial T}{\partial t}\right\},$$       (8)    

where     α = 1 − [     K     d (     t     e)/     K     s] is the Biot coefficient;     K     d (     t     e) and     K     s are the bulk moduli of the solid skeleton and solid particles of the tailings, respectively;     n is the porosity of the tailings;     t     e is the equivalent chemical reaction time at the reference temperature     T     r;     E     0(     t     e) is the confining modulus; and     β     s is the coefficient of thermal expansion of the solid components.

                 

However, Chinese researchers have also examined the impact of expansive fillers on the seepage of fissures. The primary focus of the investigation was on the comparison of the expansion pressure caused by the fillers and the mechanical parameters of the altered rock mass (Chen et al., 2006). Subsequently, Wang et al. (2010) investigated how sediment particles affect the permeability of fractured rocks. They observed that the permeability gradually increased over time and eventually reached a stable state after the transport of sediment particles resulted in re-equilibrium within the macroscopic rock fissures. In their study, Zhu et al. (2020) examined the alterations in the permeability of filled fractured rock mass under stress–permeability coupling conditions. Based on the stress–permeability coefficient curve diagram, the failure of the filling fractured rocks can be categorized into an elastic stage, an elastoplastic stage, and a final damage phase (Figure 20). Previous studies have focused on analyzing the effect of stress on the permeability of rocks with filled fissures. Based on the main idea of stress–seepage coupling, they considered the influence of stress on seepage in the case of fissures with different roughness, fissures with different inclination angles, fillers with various materials, and the selection of different kinds of rock specimens. However, the fissure channels that appear when the rock fractures under stress are the primary cause of seepage. As a result, it is also an important avenue of research to study the impact on the seepage flow of crack extension in filled fissure rocks under stress–seepage coupling.

                 

4.1.2 Analysis of seepage stresses on partially filled fissures

According to previous research, researchers here and globally have conducted numerous studies on the seepage experiments of filled fissures under stress. Nevertheless, there is a noticeable lack of research on the seepage characteristics of partially filled fissures. Therefore, Chen, Zheng, Zhang, Wang (     2019) investigated the seepage stress characteristics of incompletely filled fissures with different properties of media (brittle-plastic) to explore the corresponding seepage stress curves. Granite was selected as the filled rock sample fissure wall, made into a rectangular model; the fillers' length and height are fixed, and there are three different widths     b     0 in this experiment. The fissure width     b varies according to the variation of the filler width     b     0. The sample model is shown in Figure     21a. The generalized equation for seepage stress in incompletely filled fissures is obtained by analyzing the deformation characteristics of the filled fissures from the generalized Hooke's law:

      $$Q=\frac{n\beta b^{3}A{b}_0\Delta p{\sigma }_y}{\Delta b_0{L}^{2}E},$$       (9)    

where     Q is the flow rate;     n is the porosity of the filling medium;     β is the coefficient;     b is the width of the fissure;     A is the cross-sectional area of the specimen;     σ     y is the normal stress;     b     0 is the width of the fillers; Δ     p is the difference in pressure between the two ends of the specimen;     L is the length of the fissure;     E is the elastic modulus of the fillers; and Δ     b     0 is the transverse deformation of the fillers.

                 

A mathematical and generalized model of the seepage stress characteristics of brittle-plastic-filled fissures has also been developed based on the typical characteristics of the seepage stress in brittle-plastic-filled fissures. The segmented flow–stress relationship equation can reflect the structural change characteristics of the filled fissure before and after the peak (Chen et al., 2016):

Mathematical modeling of seepage stress in brittle-filled fissures:

            (10)    

Mathematical modeling of seepage stresses in elastic-plastic-filled fissures:

            (11)    

Mathematical modeling of seepage stress in plastic-filled fissures:

      $$Q=a{\mathrm{e}}^{-b\sigma },$$       (12)    

where     σ     1 is the normal pressure;     a and     b are coefficients, both greater than zero.

Experimental study of seepage stress characteristics of brittle-plastic incompletely filled fissures was conducted, and the results provide a technical basis for engineering problems such as seepage stress and destabilization damage of rock mass. In addition to the above, the expansion pressure of the filling material is analyzed in comparison with the mechanical parameters of the altered rock mass to explore the influence of the mechanical response of the filling material on the seepage of the fissure. Some researchers use the cusp mutation theory of microscopic structural instability as the basis to study the destabilization of half-filled fissures under seepage stress, and the method is found to be feasible (Zhang et al., 2015).

The above studies are based on single fissure filling. However, rock masses that contain fissures in nature are often intersecting and longitudinal. Hence, Zheng et al. (     2020) investigated the permeability characteristics of the rock masses by partially filling the fracture network. They modeled the fissure network based on the analytical solution of the single fissure seepage volume and the fissure network seepage principle. As shown in Figure     21b, the intersection point of a fissure is a node, and the fissure between two nodes is a line element. Then, the influence of the filling medium on the permeability characteristics of the fractured rock mass is explored through an arithmetic example. They found through testing that the presence of the filler would lead to an increase in the hydraulic-specific drop of the filled fissure, which in turn would increase the seepage flow rate. This phenomenon also increases the likelihood of localized scour erosion damage, which ultimately affects the overall strength of the fractured rock mass. Combined with the classical fissure flow theorem calculated by the cubic theorem, the mathematical model of seepage in the rock fracture network in the presence of filler in the fissure can be obtained as:

      $$\{\begin{array}{l}\mathit{A}*\left[b_{j-1}{\int }_0^{b_1}{w}_{j-x}\mathrm{d}y+b_{j-2}{\int }_{b_1}^e{u}_{j-x}\mathrm{d}y\right]+\mathit{Q}=0,\\ {H|}_{{\Gamma }_1}=h_1,\\ {H|}_{{\Gamma }_2}=h_2,\end{array}$$       (13)    

where     A is the articulation matrix of the fracture network, which is an     n ×      m order matrix;     b     j−1 is the thickness of the filler in line element     j;     b     1 is the thickness of the filler;     w     j−x is the flow rate of seepage in the filler;     b     j−2 is the thickness of the empty fissure in line element     j;     e is the width of the fissure;     u     j−x is the seepage flow rate in the empty fissure;     Q = {     Q     1,     Q     2,…,     Q     n}     T,     Q     i is the source term of node     i, and     i takes the values of 1,2,3,…,     n;     Γ     1 and     Γ     2 are the boundaries of the known heads, respectively; and     h     1 and     h     2 are the heads on the corresponding boundaries, respectively. By solving Equation (     13) through Matlab programming, the head at each node can be obtained.

In summary, seepage pressure plays a role in promoting the growth of filled fractures. The observation of the crack distribution on specimen surfaces revealed that the filler tended to slip into these cracks. The mechanical behavior of fillers significantly impacts the deformation of fractured rock masses. Both tensile and shear effects contribute to an increase in fissure permeability. Furthermore, when fillers are subjected to external forces, liquefaction and plasticization effects also lead to a significant increase in fissure permeability. Studies to date have mainly explored the macroscopic seepage characteristics of filled fractures, with limited attention paid to microscopic perspectives. Hence, research on the seepage testing of filled-fracture networks remains in the early stages. The complexity of fracture networks poses difficulties in test design and equipment selection, warranting further strengthening of research in this field.

4.2 Influence of shear on the flow characteristics of filled fissures

Single fissure seepage, as the basis of seepage in fractured rocks, has been extensively studied by researchers at home and abroad, and many models have been proposed. However, these models basically study the coupling between normal stress and seepage, and there are few studies on shear stress and even fewer studies on shear deformation. Shear expansive deformation is the deformation of a fissure expansion resulting from shear displacement along a nonplanar fissure surface, leading to perturbation of the fissure geometry and characteristics, including changes in the contact ratio, roughness, and jointing. Shearing of pre-existing fissure surfaces produces bulges and inter-sliding, leading to surface disruption and reorganization of the bulges so that fluids can pass through these expanding or re-fractured permeable channels (Ai-Yaarubi et al., 2004; Baghbanan et al., 2008; Detwiler & Morris, 2014; Zhou et al., 2018).

Therefore, it has been shown that shear stress has a significant effect on fissure seepage through studies on the coupling of shear stress and seepage (Jiang et al., 2007; Koyama et al., 2008; Liu et al., 2003). Xue et al. (2007) established a coupling mechanism between stress–strain and seepage in rock fissures by considering only the shear stress and shear deformation of rock fissures. The changes and interrelationships of the shear deformation, openness, hydraulic conductivity, seepage field, and stress field of the fissure are studied. Xu et al. (2009) conducted experimental studies on the permeability properties of shear fissures under fissure water pressure and different peripheral pressures. They derived an equation for calculating the permeability coefficient that takes into account the fissure water pressure. Xiong et al. (2010) validated the reliability of using the numerical method of the three-dimensional Navier–Stokes equations to model the fluid flow state within the fissure. During the coupled shear–seepage test, the hydraulic conductivity coefficient of the fissure showed three distinct stages of change. Wang et al. (2004) performed shear–permeability coupling tests and determined the variation pattern of hydraulic opening during the shear test. They found that the fractal correction formula closely corresponded to the experimental test values. Furthermore, a comprehensive understanding of the seepage characteristics of the nodal surface was achieved through the coupling of the contact area of the nodal surface, the fractal dimension of the nodal surface, and the nodal opening. Fan et al. (2020) investigated the seepage law of marble shear fissures. It was found that the local tensile damage characteristics produced in the process of shearing play a certain role in hindering the seepage of shear fissures. The effect of shear deformation on fracture permeability has a more complex and variable relationship. The results on the permeability of fissure joints rely on the magnitude of shear displacement, rough surface shear damage, and the shape of the joint surface (Li, Wang, et al., 2021; Sheng et al., 2021; Zhao, 2021).

It can be found from the above studies that research on seepage through fissures is extensive. Admittedly, natural rock fissures are generally filled with filling mediums, and yet, research results on the effect of shear on seepage through filled fissures are rare. Liu et al. (2002) carried out shear–seepage experiments on the controlled displacement of sand-filling fissures. It was revealed that the flow rate during the shear displacement of rock fissures increased with the increase of shear deformation combined with the physical properties of the filled sand. Cheng et al. (2018) conducted a study using the discrete unit program PFC2D to analyze how rock joints with different roughness levels and filler thicknesses affect mechanical properties and permeability evolution during shear. The researchers measured the pattern of permeability changes during shear and observed that the permeability of filled rock joints increased as the thickness ratio and JRC increased. Xu, Wu, et al. (2019) carried out experimental studies on the shear–permeability coupling characteristics of similar materials in sandstones with structural surfaces under different filling degrees by utilizing a self-developed coal rock shear–permeability coupling test device. Figure 22 illustrates the flow–shear displacement curves for the simulated joints at various filling degrees. It is evident that the fracture width expands due to shear, resulting in the enlargement of the seepage channel. Consequently, the flow rate shows a slight increase in the later stage. Du et al. (2023) investigated the influence of fracture fillings on mechanical characteristics, fracture behavior, and permeability evolution during seepage stress damage in jointed rock masses. It is mainly dominated by shear in the peak damage phase, which also provides channels for seepage. Hence, the permeability increases abruptly in this phase. The decrease in permeability after the peak is due to the closure of the shear channel by the presence of the filling material (Figure 23). As a result, the evolution of fracture opening during shear induces stress concentration on the structural surface, which in turn leads to a notable alteration in the flow rate. In addition, the influence of the filling body on the permeability of a rock mass with fractures is closely related to the cracking behavior of the fractures.

               

               

               

In summary, under the action of shear stress, the gap width is the main factor that affects the seepage of rock masses, regardless of whether the fissure is filled or unfilled. Consideration of seepage characteristics under shear stress, along with coupled stress–strain analysis at low stress levels, is crucial for engineering safety. Currently, the shear–seepage coupling test studies on rock masses with structural surfaces mostly focus on cases with single fissures and no filler. Also, studies on fracture networks, filled fissures, and other complex fissures are relatively limited. Future research should consider the influence of physical factors such as the types of filling material, filling thickness, and filling particle size. Then, we can establish the fracture network-filled model by considering these factors. Finally, combining macroscopic and microscopic perspectives will enable a deeper exploration of the mechanisms of rock mass seepage and rupture.