In the natural environment, rocks are composed of diverse mineral particles and cementing materials that are bonded together. And their mechanical properties are mainly affected by the friction between particles. Considering that sand particles can provide a frictional force, cement mortar was chosen as the rock-like material for this experiment.

Sedimentary rocks, which account for approximately two-thirds of the Earth’s land area, are widely distributed on the Earth’s surface. To create composite rock-like materials, limestone and sandstone were chosen as the original rocks, which exhibit distinct differences in physical properties and are commonly found in sedimentary rocks. The quartz sand was used as an aggregate of rock-like material. To ensure a smooth specimen surface, the size gradation distribution of the selected quartz sand particles is presented in Table 1. P.O42.5 Portland cement, silica fume, and iron powder were used as limestone-like materials. The purpose of using silica fume was to lower the peak temperature of the material’s curing reaction and prevent cracking. The purpose of selecting iron powder was to improve the density of the rock. To differentiate the sandstone in terms of mechanical properties and visual appearance, P.W.32.5 Portland cement was used in its production. Additionally, a small amount of water reducer and defoamer was added to enhance the fluidity of the material and minimize the formation of pores. Table 2 provides the mass ratios of rock-like material. Table 3 presents the physico-mechanical parameters of cement mortar, which are close to that of natural rocks (Nazir et al., 2013; Cai et al., 2022). Therefore, cement mortar is considered an ideal rock-like material for laboratory experiments.

During the solidification process, cement tends to form a thin layer on the surface (Wang et al., 2018), which is prone to cracking under compression. To ensure accurate observation of real cracks in the specimens, this study first, poured cubic specimens (height × width × thickness = 150 × 150 × 150 mm) and then obtained cylindrical specimens by drilling, as illustrated in Figure 1A For cubic specimen casting, custom molds were created using plastic, steel sheets (15 mm length, 0.5 mm thickness), and spacer boards. Vibrations were applied to the mold during pouring to eliminate trapped air from the slurry. After 30 min, the spacer board was removed to ensure sufficient strength at the contact surface between the two types of rock materials. After 8 h, the steel sheets were removed when the specimens reached the initial setting, resulting in through-opened fissures. After 24 h, the solidified specimens were de-molded and placed in a constant temperature and humidity curing chamber (temperature set to 20°C ± 2°C and humidity set to 85% ± 5%) for 28 days before conducting cylinder specimen drilling.

Following the guidelines of the International Society for Rock Mechanics (ISRM), cylindrical specimens were cut and ground to a height of 100 mm and a diameter of 50 mm. To facilitate the analysis of the effects of fissure angle, position, and confining pressure on the specimens and to prevent the mechanical failure patterns of the specimens from being universally applicable due to excessively long fissure lengths, the fissure length was fixed at 15 mm. Five fissure angles (α) were considered: 0°, 30°, 45°, 60°, and 90°. Three different fissure positions were established: ①at the center of the limestone, ②at the center of the contact interface (in this study, the term “interface” is used to describe the contact interface parameters of limestone and sandstone), and ③at the center of the sandstone, as shown in Figure 1B Four different confining pressures (σ ₃) were applied: 0, 5, 10, and 15 MPa. Considering the uniaxial compressive strength of the simulated sandstone as 40 MPa, and ensuring that the deformation of the specimen remains in the elastic stage during the confining pressure process (typically, the ratio of the threshold stress for sandstone fracturing to the peak stress is approximately 0.5), the maximum value of confining pressure was set to 15 MPa.

For clarity, the specimen numbering is designated as fissure position (FL, FI, FS, C) - fissure angle (α) - confining pressure (σ ₃), where FL, FI, FS, and C represent the fissure positions in limestone, interface, sandstone, and the complete composite rock specimen, respectively. For instance, C-5 denotes the complete composite rock specimen under a confining pressure of 5 MPa. FL-45-15 signifies a composite rock specimen with a fissure angle of 45° located in limestone under a confining pressure of 15 MPa. The specific specimen numbering is outlined in Table 4. To obtain more universally applicable experimental data and avoid significant variability in individual specimens, each set of specimens underwent four tests. After removing specimens with noticeable variability, the average of the remaining specimen data was taken as the result for each set of specimens. This approach helps reduce the impact of individual outlier test data on the overall results, enhancing the reliability of the experimental outcomes.

Triaxial compression strength (TCS) test was processed by HYAS-1000C electrohydraulic servo-controlled testing equipment at the geotechnical mechanics and engineering research center at Yangtze University in China, as shown in Figure 2. In the experiment, the confining pressure was applied using hydraulic oil, while the axial stress was controlled by axial displacement. Before applying the confining pressure, the specimen was placed between the upper and lower platens and encapsulated with heat-shrinkable tubing to prevent hydraulic oil from infiltrating the rock and affecting the test results. The axial and radial strains of the specimen (at the middle section) were simultaneously measured using axial and radial extensometers, with a measurement range of 0–4 mm and a reading accuracy of ±1% at room temperature. Data collection was performed directly by a computer. The test was conducted in two steps as follows: (a) The confining pressure was increased to the target value at a rate of 0.5 MPa/min, then the axial compression was applied to the specimen at a rate of 1 mm/min until reaching an initial axial stress of 2 KN. Additional axial stress (also called deviator stress) was applied to the specimen at a rate of 0.2 mm/min until it experienced unstable failure (Yu et al., 2021b).

To better investigate the mechanical properties of fissured composite rock samples, uniaxial compression tests were initially conducted on individual sandstone, individual limestone, and intact composite rock, and the stress-strain curves and equivalent mechanical parameters are obtained, as shown in Figure 3 and Table 5.

Figure 3A illustrates notable distinctions among the three specimens, particularly in the pre-peak stage. The limestone’s compaction phase is relatively inconspicuous, featuring an extended elastic phase. In contrast, composite rock and sandstone show significant compaction stages, with sandstone exhibiting a shorter elastic phase and a slight stress reduction just before the peak. This suggests high compaction in limestone and greater porosity in sandstone, making it more prone to macroscopic cracks before complete failure. Despite these differences, all uniaxial stress-strain curves share similar post-peak characteristics, marked by a rapid stress decline towards residual strength and a swift loss of bearing capacity.

Statistical analysis of experimental data (Figure 3B; Table 5) reveals limestone’s superior strength, surpassing composite rock and sandstone by 86.55% and 96.41%, respectively. Additionally, its elastic modulus exceeds that of composite rock by 47.82% and sandstone by 59.09%. Regarding deformation characteristics, limestone has the lowest peak strain, 12.82% less than sandstone, while the peak strain of composite rock falls between the two.

Figure 3A depicts the failure characteristics of three specimen types. Limestone and composite rock experience tensile failure along the axial compression direction, while sandstone undergoes a tensile-shear mixed failure. Deformation upon sandstone failure is notably larger than in limestone, marked by evident surface spalling. Due to lower sandstone strength, composite rock failure initiates in the sandstone portion, forming macroscopic cracks that propagate to the contact interface, penetrating through the limestone, where spalling primarily occurs.

In summary, there exists a significant strength difference between limestone and sandstone. Sandstone, being relatively softer in lithology compared to the harder limestone, exhibits larger deformations under compression. The strength of the composite rock is determined by the weaker sandstone portion, yet its deformation is constrained by the stronger limestone portion.

In the compression failure modes of rocks, cracks can be classified into three forms based on the angle between the generated cracks and fissures, as well as the scratches on the fractured surface (Lee and Rathnaweera, 2016; Lin et al., 2020; Yu et al., 2021b): (1) Axial splitting, which refers to tensile cracks formed under compression, generally parallel to the direction of maximum principal stress, mostly with an angle greater than 90° to the fissures. The crack fracture edges are neat, and the crack fractured surface is smooth, with only minor undulations in certain areas, as indicated by the red region in Figure 4A; (2) Shear cracks, typically composed of the main shear crack and secondary tensile cracks on the specimen surface. Secondary tensile cracks originate from the main shear crack and extend, with the main shear crack coplanar or less than 90° to the fissures. Most observed shear cracks exhibit evident shear dilation characteristics, with serrated fracture edges and a rough fractured surface. In the limestone portion, there is also a strong frictional effect resulting in white powder. Additionally, at lower confining pressures, shear cracks terminate at the end faces, while at higher confining pressures, shear cracks terminate at the specimen’s side, as illustrated by the blue region in Figure 4B; (3) Tensile-shear mixed cracks, observed only in some specimens. These cracks typically start as tensile cracks extending towards the ends and then shift to extend to the specimen’s side, forming shear cracks, as shown in Figure 4C.

To visually illustrate the failure characteristics of the specimens more clearly, clear crack trajectories were plotted. The red and blue lines represent the generated main tensile cracks and main shear cracks, respectively. The black fine lines on the specimen surface depict secondary tensile cracks, which do not influence the internal structure of the specimen.

Figure 5 presents the failure modes of composite rock specimens with different fissure angles and positions under a confining pressure of 5 MPa. By comparing various aspects, the characteristics of the cracks and their aggregation forms are summarized.

Under different fissure angles: (1) When α = 0°, regardless of the rock layer in which the fissure is located, tensile cracks always initiate at the middle of the fissure and extend towards the specimen ends. This is because the fissure is perpendicular to the direction of the principal stress, and the inner surface of the fissure is in a bending state. With the increase in axial load, tensile cracks are prone to occur in the middle of the fissure. (2) When α = 30°–60°, as the region with the most concentrated stress, wing cracks or anti-wing cracks are likely to form at the tips of the fissure, resulting in similar failure modes. (3) When α = 90°, large compressive forces leading to tensile cracks are less likely to occur in the fissure. Due to the strength difference between the two types of rocks, shear cracks first initiate and propagate in the sandstone, forming a wedge shape that causes limestone to undergo splitting failure. The failure mode of the specimen mainly depends on the fissure positions.

Under different fissure positions: (1) When the fissure is in limestone, some specimens may develop a small number of shear cracks, but overall, tensile failure predominates. Under α = 0°–60°, tensile cracks generated at the lower part and the lower tip of the fissure easily propagate through the contact interface towards the sandstone end, causing overall failure of the composite rock. However, when α = 0° and 30°, due to the lower strength of the specimen, tensile cracks do not extend to the sandstone end but continue to form shear cracks in the sandstone under the action of axial compression and confining pressure. (2) When the fissure is at the contact interface position, combined with Figure 6, it is observed that in the sandstone part, shear failure leads to the formation of wedge-shaped blocks with distinct signs of sliding friction on the surface. In the limestone part, tensile failure results in a smooth and neat fracture surface, and the failure mode of the specimen is not affected by the fissure angle, exhibiting “λ-shaped failure.” This is because the strength of limestone is greater than that of sandstone, and shear cracks generated at both ends of the fissure first extend towards the sandstone end. After forming a wedge-shaped body in the sandstone portion, it continues to be compressed by axial pressure, ultimately leading to tensile failure in limestone. (3) When the fissure is in sandstone, the specimen mainly experiences localized shear failure in the sandstone. This is because the fissure further reduces the strength of the sandstone, and anti-wing shear cracks developed at the ends of the fissure receive sufficient development under lower stress levels. The cracks extend towards the sandstone end and side. Meanwhile, tensile cracks generated by the fissure extend towards the limestone but do not penetrate the limestone portion. (4) Surface sliding and spalling: Surface sliding and spalling mainly occur at the sandstone end and the tip of the fissure. This is primarily due to the sandstone portion and the tip of the fissure being more susceptible to the compressive squeezing effect of confining pressure, resulting in localized substantial shear dilation deformation.

Figure 7 illustrates the failure modes of composite rock specimens under different confining pressures and fissure positions at α = 45°. Under uniaxial compression conditions, the composite rock specimens exhibit conspicuous splitting failure caused by a tensile crack extending along the direction of stress loading (cracks generated can be merged into one). With increasing confining pressure, the following observations are made: (1) The failure modes of all specimens gradually transition from tensile failure to shear failure. This is attributed to high confining pressure exerting closure forces on nearly longitudinal cracks, inhibiting the extension of wing-shaped tensile cracks. Additionally, the shear stress acting on the fissure tip becomes more significant, leading to a purer shear failure of the specimen. (2) All specimens experience overall failure. Combined with Figure 8, it is observed that due to the gradual enhancement of shear stress, shear cracks appear at the tips of the fissure in limestone, and the frictional sliding traces on the internal crack fracture surface become more apparent. This ultimately leads to the formation of wedge-shaped bodies in limestone, continuing to be compressed and resulting in overall specimen failure. When the fissure is at the contact interface and in the sandstone, co-planar shear cracks occur at both fissure tips. The specimen slides along the fissure to some extent, and the shear stress developed in the sandstone can extend shear cracks into the limestone, causing shear failure in the limestone portion as well. (3) With the increase in confining pressure, there is an increase in surface spalling. This is attributed to higher confining pressure leading to more effective compaction of the fissure. In this state, the interlocking effect at the fissure tip becomes increasingly pronounced, promoting the extensive development of secondary cracks and resulting in a greater occurrence of local spalling phenomena.

In summary, the geometric distribution of a single fissure has a significant impact on the failure patterns of composite rocks under confining pressure conditions. The analysis of the failure characteristics of each specimen is summarized in Tables 6, 7. From Table 6, it is evident that maintaining a constant confining pressure (σ ₃ = 5 MPa), fissure in limestone is prone to generate tensile cracks, leading to the overall failure of the specimen. On the other hand, the fissure tips in sandstone primarily produce shear cracks that are challenging to extend through the contact interface into the limestone, resulting in local failure of the specimen. In this case, the failure mode of the composite rock is dominated by the fissure position, while changes in the fissure angle primarily influence the initiation mode of the specimens. From Table 7, under uniaxial compression conditions, the specimens primarily experience axial splitting failure. With the increase in confining pressure, the failure mode of the specimens gradually transitions from tensile failure to overall shear failure. The influence of the fissure position on the failure of the composite rock becomes less pronounced, indicating a gradual shift in the failure mode of the composite rock to be dominated by the confining pressure.

Under 5 MPa confining pressure, the full stress-strain curves of specimens with different fissure angles and positions are depicted in Figure 9. The volumetric strain reflects the overall volume change of the specimen during the loading process and is calculated as follows: ε _V = ε ₁ + 2ε ₃. It can be observed that, with the fissure position held constant, the axial stress-strain trends of all specimens are nearly identical. In the pre-peak stage, as the fissure angle increases, the linear elastic stage of the specimen significantly extends, and the peak stress also rises, gradually approaching that of the intact specimen. However, the plastic yield stage of each specimen is more pronounced compared to the intact specimen. This indicates that an increase in the fissure angle can enhance the strength of the specimen, but the presence of fissure still increases the accumulated damage within the specimen, leading to an elongation of the plastic stage experienced by the specimen before complete failure. In the post-peak stage, due to the constraint of confining pressure, all specimens do not exhibit significant brittle failure characteristics.

The volumetric change trends of all specimens show a pattern of initial contraction followed by expansion. In terms of contraction, the peak volumetric strain of each specimen increases with the rise in fissure angle. This indicates that as the linear elastic stage of the specimen becomes longer with an increasing fissure angle, the ability of the specimen to contract becomes more pronounced. However, the intact rock specimen, without the presence of fissures, does not exhibit significant compression and, therefore, does not possess the strongest contraction ability. In terms of expansion, under the same fissure angle, when the fissure is in the sandstone, the specimen exhibits the weakest expansion ability. This is because the FS series specimens experience localized failure in the sandstone during destruction, resulting in a relatively insignificant overall expansion of the specimen. Therefore, at lower confining pressures (σ ₃ = 5 MPa), the contraction of the specimen is dominated by the fissure angle, while the expansion is dominated by the fissure position.

The full stress-strain curves of various specimens under different confining pressure levels and fissure positions are illustrated in Figure 10. Under uniaxial compression conditions, each specimen exhibits a prolonged compaction stage, sharp peaks appear at the peak stress, followed by a rapid decline after the peak. When the confining pressure is 5 and 10 MPa, the composite rock specimen with a fissure in the limestone part exhibits a stress drop during the loading process, accompanied by a sudden increase in lateral strain, indicating the initiation of macroscopic cracks. The stress-strain curves of the other two types of fissure specimens show similar behavior at this point, with the deformation characteristics of the specimen dominated by the fissure position. As the confining pressure increases to 15 MPa, the compaction stage disappears gradually. The elastic stage and yield stage before the peak becomes significantly extended, and both peak stress and peak strain increase accordingly. This is because, before axial compression, the micro-pores and micro-cracks inside the specimen have been compacted and closed by the confining pressure. After the elastic stage, the macroscopic cracks generated are constrained by the confining pressure and their crack fractured surface interlock and friction with each other. This results in a slow increase in axial bearing capacity, leading to the formation of a plastic flow state characterized by strain hardening during the yield stage. This significantly enhances the strength and ductility of the specimen. At this point, the influence of the fissure position on the mechanical deformation characteristics of the rock specimen diminishes, gradually being dominated by the confining pressure.

Under uniaxial compression conditions, the maximum values of volumetric contraction for all specimens are relatively small and similar. With the increase in confining pressure, the maximum values of volumetric contraction for most specimens become closer, showing an overall trend of slow growth at high confining pressures. In terms of volumetric expansion, regardless of the magnitude of the confining pressure, the degree of volumetric expansion in the composite rock specimens with a fissure in the sandstone portion is the smallest. Even after complete failure under high confining pressure, the overall volume remains contracted. The reason is that when the fissure is in the sandstone, the damage to the limestone part is minimal, resulting in the smallest overall expansion of the specimen. Therefore, the volumetric contraction of the specimen is dominated by confining pressure, while volumetric expansion is dominated by the fissure position.