Failure characteristics and control technology of surrounding rock during roadway passing collapse column Authors: Ce Jia, Chaojun Fan, Sheng Li, Xiaoshuai Liu, Mingkun Luo Published: 2024 ABSTRACT AbstractThe collapse column may result in failure of the surrounding rock during roadway excavation. It is of considerable importance to elucidate the failure characteristics and develop the corresponding control technologies for the surrounding rock. A discrete fracture mesh (DFN) and a strain-softening constitutive model are employed to simulate the properties of a collapse column. The stress zoning model was developed to characterize the rock surrounding the collapse column. Additionally, the zoned support technology was proposed to stabilize the roadway surrounding the rock. From the center outward, the results indicate that the collapse column can be individually divided into four zones: an internal impact zone of the collapse column, a stress decreasing zone, a stress increasing zone, and an initial stress zone. Upon the visibility of the collapse column, the tensile failure occurs in the roof, leading to a decrease in stress in front of the roadway and an increase in stress on both sides of the collapse column. The collapse amplifies the range of plastic failure by inducing an overlap between the roof stress and the stress elevation zones. The impact zone of the collapse column was reinforced by means of an integrated support system consisting of “pre-grouting, bolt cable mesh, and steel shed”. The “bolt cable mesh” provides integrated support beyond the area affected by the collapse column. Following a 20d excavation period, the roof and the two sides were stabilized, with an effectively controlling for the surrounding rock.Keywords: Coal mine roadway, Collapse column, Failure characteristics, Surrounding rock of roadway, Control technologySubject terms: Civil engineering, Coal Introduction According to statistical data, a total of 2800 karst collapse columns have been recorded across 22 provinces in China1. In addition to impeding the safe and efficient production of coal mines, these collapse columns also lead to a reduction in recoverable reserves and significantly impact coal mining operations2–4. The collapse column poses several challenges, including the potential occurrence of water inrush accidents resulting from the collapse column5,6. Moreover the fractured zone of the collapse column undergoes continuous development and induces destabilization in the surrounding rock due to mining activities. Additionally, the stress concentration zones are formed within the collapse column due to its own weight7,8. For efficient and secure excavation of roadway, it is crucial to investigate the failure characteristics and control technology of the surrounding rock during development roadways through collapse column formation. Water disasters are likely to occur during roadway excavation when encountering a collapse column9–11. Some scholars have carried out theoretical studies on the collapse mechanism of collapse column12–15, such as the thick wall model for the circular hole16, the elliptical cross-section model17, and the mechanical model of “four bands” for collapse column18 to investigate water-related incidents associated with collapse column. The collapse column leads to changes in the initial rock stress field. Therefore, it is significant to grasp the stress evolution law and failure characteristics of the surrounding rock during excavating through the collapse column. Xu19 investigated the stress distribution pattern surrounding the collapse column by employing the elastic-plastic theory and numerical simulation methodologies. Ma20 established a mechanical model of the collapse column within the coal seam and used the FLAC3D numerical software to simulate the stress and failure law of the surrounding rock during coal mining through the collapse column. Based on this premise, Zhang5 investigate the existence of pores inside the collapse column and employed the FLAC3D software to establish a random pore collapse column model. Moreover, the stress distribution and failure law of mining coal seam through the surrounding rock of the collapse column were obtained. Zuo21 analyzed the stress distribution pattern of the surrounding rock in the fracture zone around the collapse column using two-dimensional discontinuity (DDA) software. The aforementioned studies have successfully simulated the effects of collapse columns on the surrounding rock. However, the studies have not comprehensively addressed the impact of internal fractures within the collapse columns on the control and behavior of the surrounding rock. Based on numerical calculations, scholars22,23 investigated the rupture law of the surrounding rock, and the evolution law of the support pressure at the working face pushed through the collapse column based on the similarity theory. The presence of the collapse column complicates the tectonic stress field in the surrounding rock, leading to the broken surrounding rock and the challenges in providing the adequate support. In the direction of enhancing the strength of the surrounding rock, Zhang24 introduced a self-stress grouting reinforcement technique. Kang25 developed a composite grouting material and proposed a pre-grouting approach prior to the excavation working face. Additionally, Ni26 put forward a grouting bolt support technology. The reinforcement of support in the surrounding rock was addressed by Jian27 through proposing an integrated support system consisting of “bolt, W steel belt” + “anchor cable, anchor cable beam” + “I-steel shed”. Zhang28 effectively solved the issue of the cable being unable to establish suspension anchorage within the failure zone of the collapse column by employing an encrypted arch shed support. Zuo29 proposed the synergistic support method involving the utilization of “appropriately encrypted anchor rods, large-diameter high-strength anchor ropes, reinforcing mesh spraying layer, and lagging grouting.” The aforementioned researchers have conducted studies on the failure characteristics and control methodologies of the collapse column. However, the simulation of the collapse column did not account for the loose and fissure characteristics, failed to specify the stress zoning of the surrounding rock around the collapse column, and neglected to propose the surrounding rock zoning support technology. This paper utilizes the internal looseness and fracture properties of the simulated collapse column using the DFN and the strain-softening constitutive model. Additionally, a stress zoning model is provided for the surrounding area of the collapse column. This study investigates the stress evolution law and failure characteristics of the surrounding rock in relation to the collapse column. Consequently, the influence zone of collapse column was determined and the zonal support measures were proposed and verified by on-site engineering practice. Failure characterization of surrounding rock during roadway passing collapse column Research background The Zhangcun Coal Mine is located in Changzhi City, Shanxi Province, and mainly mines the No.3 coal seam, which possesses an average thickness of 5.69 m. The burial depth of the mining operation is 500–517 m, and it utilizes a fully automated top coal caving mining technique. The immediate roof of the roadway consists of mudstone with an average thickness of 2.62 m; The basic roof of the roadway is composed of sandstone with an average thickness of 9.22 m; The immediate floor of the roadway is made up of mudstone with an average thickness of 3.46 m; The basic floor of the roadway comprises fine-grained sandstone with an average thickness of 5.67 m, as shown in Fig. 1(a). The 480-material roadway refers to the + 480 m mining level development roadway that is excavated along the roof within the coal seam. This roadway is constructed with a rectangular section, measuring 5.6 m in length and 3.7 m in height. The 480-material roadway and 480-belt roadway are arranged parallel, with a separation distance of 55 m. In front of the 480-material roadway, there is a collapse column labeled as X20, which intersects with section III of the 480-material roadway, as shown in Fig. 1(b). Notably, no water discharge is observed from the drill hole during the identification process of the collapse column, indicating an absence of potential water conduction within the collapse column. The presence of a collapse column has a detrimental impact on the stability of the surrounding rock and is not conducive to support for the roadway. In the collapsed column area, this study employed a core drilling method to ensure the integrity of the obtained rock samples. Given the relatively soft texture of the rock samples in the collapsed column fractured zone, the drilling speed was reduced to minimize damage to the samples during the extraction process. Subsequently, to prepare standard-compliant specimens for subsequent strength testing, the YSXQ-II diamond wire CNC cutting machine was used for precise cutting of the rock samples. Geological conditions of 480-material roadway. Modeling process and methods The engineering background for this study was the 480-material roadway in Zhangcun coal mine. A three-dimensional geometric model was constructed, with dimensions of 300 m (length) × 240 m (width) × 27 m (height), as shown in Fig. 2. The collapse column is situated directly in front of the roadway, and the length of the study roadway is 240 m. The roadway is divided into two sections based on the surrounding rock characteristics. The first section, ranging from 0 to 55 m and 185–240 m, is classified as normal surrounding rock. The second section, from 55 to 185 m, is categorized as the geological abnormality zone associated with the collapse column. The lower boundary of the model was set as a fixed constraint, while normal displacement constraints were applied to the surrounding boundaries. Additionally, the upper boundary was set as the gravity load of the overlying rock. 3D model. The fractures in the collapse column conform to a random distribution, and the Monte-Carlo method is employed to implement the fracture stochastic simulation process. The features of the Discrete Fracture Network (DFN) were characterized using the FLAC3D software, which involved determining the size, location, and orientation of the fractures. It is assumed that the distribution of fracture sizes follows a negative power-law distribution30. Where n (l) represents the length of the fracture size, m; α represents the density term; a represents the power function index, which takes the value of 3 to 4, employed to control the ratio between large and small-sized fracture. As the “a” increases, so does the size of the fracture. Integrate Eq. (1) to obtain the maximum and minimum dimensions of the fissure: Where Lmin represents the minimum side length of the cube, m; Lmax represents the maximum side length of the cube, m. The spatial position of the DFN fissure inside the collapse column conforms to the most widely used Poisson process. In this paper, it is assumed that the central point of the 3D fissure obeys the Poisson point process31. Each individual fissure is considered independent from others, and a random uniform distribution function is used to obtain the coordinates of the fissure centers (x, y, z). Fracture azimuths are vectors, containing dip and inclination, and are assumed to follow the Fisher distribution32. Where θ represents the fissure dip angle, °, 0 ≤ θ ≤ 90°; Φ represents the fissure dip direction, °, 0 ≤ Φ ≤ 360°; k represents the coefficient of discretization, and the larger the value of k is the more concentrated the data constraints are. The fracture mass density reaches a threshold fracture generation termination33: Where: dm (lc) represents the cumulative fissure surface area per unit volume, m2/m3; l2 represents the fissure area, m2; L represents the dimension length, m. The attribute parameters of DFN were set according to Eq. (1) to (4), whose parameter ranges cover the literature34–36 and are listed in Table 1. Table 1DFN attribute parameter settings.ParameterDistribution functionValueFracture sizePower-law distributiona is 3.5, Lmin is 0.5 m, Lmax is 3 mFracture orientationFisher distributionθ is 60°, Φ is 250°, K is 5Fracture mass density0.33 DFN attribute parameter settings. This study establishes a Discrete Fracture Network (DFN) within a 55–185 m section of a 480-material roadway to simulate the impact of random fissures within the collapse column on the support system. The method of DFN construction entails establishing the size, location, and orientation, as illustrated in Fig. 3. The internal fracture reconstruction process of collapse column. The strength parameters in the crushed zone of the collapse column exhibited nonlinear variations, and the change in strength deterioration became more pronounced with the increase in stress. Therefore, a strain-softened constitutive model was employed to simulate the internal loosening characteristics of the collapse column. The strain softening constitutive model is based on an improved Moore-Coulomb constitutive model with uncorrelated shear and correlated tensorial flow rules. This principal structure accounts for the softening of the material cohesion, the angle of internal friction, and the tensile strength upon reaching the yield state.” After the model material reaches the yield state, both elastic and plastic strains are generated, as shown in Fig. 4, at which time the strain is: Stress-strain relationship. Where ε represents the total strain; εe represents the elastic strain; εp represents the plastic strain. The plastic shear strain in the yield failure of a collapse column as a function of cohesion, the angle of internal friction as an independent variable is defined as37: Where , , j = 1, 3 represents the plastic shear principal strain increment. Plastic tensile strain is defined as a function independent variable of tensile strength: Where represents the plastic strain increment in the direction of the principal stress. The model material yield failure criterion is38: Where σ1 represents the maximum principal stress, MPa; σ3 represents the minimum principal stress, MPa; fs=0 represents the shear failure; ft=0 represents the tensile failure; c represents the cohesive force, MPa; φ represents the angle of internal friction, °; Nφ = 1 + sinφ/1-sinφ. The FLAC3D software was used to simulate the uniaxial compression of the specimen within the crushed zone of the collapse column. This study aims to ascertain the softening behavior of cohesion, friction angle, and tensile strength in specimens located within the crushed zone of the collapse column. The simulation parameters were calibrated and depicted in Fig. 5. Softening functions for cohesion, friction angle, and tensile strength were determined and listed in Table 2. Calibration of numerical simulation parameters. Table 2Strain softening intrinsic parameter settings.Shear strainCohesion /MPaAngle of internal friction /(°)Tensile strainTensile strength /MPa0.000 C P 0.000 Tp 0.005CP/5-2.50.005Tp-2.50.0100-50.010Tp-50.0500-50.050Tp-5 Strain softening intrinsic parameter settings. The Mohr-Coulomb constitutive model was employed to simulate the remaining formations. To approach the actual parameters in the field, the physical and mechanical parameters of each formation were strength discounting based on laboratory measurement data39,40, as listed in Table 3. Table 3Numerical simulation of coal and rock mechanical parameters.Rock formationDensity /(kg·cm-3)Bulk modulus/(GPa)Shear modulus/(GPa)Cohesion/(MPa)Angle of internal friction/(°)Tensile strength/(MPa)Sandstone26505.55.24.66341.84Mudstone24502.862.172.9301.6Coal14003.12.41.98250.41Collapse column17002.72.63.1220.7 Numerical simulation of coal and rock mechanical parameters. Analysis and discussion on the stress evolution law of surrounding rock (1) Stress distribution pattern of the original rock. The vertical stresses inside the collapse column, as depicted in Fig. 6, are comparatively lower than those observed at the peripheral boundaries of the collapse column. Notably, high stress levels are evident at the periphery of the collapse column. If the ultimate strength of the rock mass is exceeded, the peripheral boundary of the collapse column is prone to destabilization failure. The stress changes from the interior of the collapse column to the peripheral boundary, exhibiting a gradual decrease followed by an increase, creating a stress decrease zone and an increase zone. As the distance of the peripheral rock body from the boundary of the collapse column increases, the vertical stress gradually decreases and eventually reaches a stable state. The zoning of the collapse column consists of four distinct regions: I - the internal impact zone of the collapsed column, II- the stress-decreasing zone, III- the stress-increasing zone, and IV - the initial stress zone, in sequential sequence. The numerical simulation results presented in this paper are consistent with the pattern of numerical simulations in the references41. Distribution law of original rock stress. (2) Analysis of the stress evolution law of the surrounding rock before exposing the collapse column. Figure 7 illustrates the formation of an elliptical high-stress zone around the collapse column. During the roadway excavation, the stresses of the roof and the excavation working face are elevated and eventually coupled with the elliptical high-stress zone. If the ultimate strength of the collapse column is exceeded, the initial failure occurs at the top of the collapse column and the ring boundary. When the roadway was excavated for 30 m, the stress in front of the excavation working face increased from 17 MPa to 18 MPa, an increase of 6%. When the roadway was excavated for 50 m, the stress in front of the excavation working face suddenly increased to 20 MPa, an increase of 11.11%. At this time, the distance between the roadway and the collapse column is getting closer and closer, while the inside of the collapse column is loose. Consequently, the collapse column fails to provide support to the roof, resulting in an increased stress in front of the excavation working face. Vertical stress distribution in the surrounding rock before exposing the collapse column. (2) Analysis of the stress evolution law of the surrounding rock during the exposure of collapse column. Figure 8 demonstrates the superimposed coupling of the roof stress with the elliptical high-stress zone when the roadway was excavated for 60 m, and the stresses on both sides of the collapse column are strengthened. The stress in front of the excavation working face experienced a sudden reduction from 20 MPa to 14 MPa, representing a decrease of 30%. The interior of the collapse column is loose and does not support the roof of the roadway, resulting in significant failure and a decrease in stress. The excavation of the roadway extended for 120 m, and the collapse column was exposed for 70 m. The stress of the roof increased from 20 MPa to 22 MPa, an increase of 9%. Furthermore, there was a secondary increase in stress observed on both sides of the collapse column. The stress in front of the excavation working face increased from 14 MPa to 16 MPa, an increase of 14.3%. The minor failure range of the roof at this stage is attributed to the low stress in the center of the collapse column, and the stress concentration is enhanced. When the collapse column is exposed for 130 m, the stress in the roof increases from 22 MPa to 23 MPa, an increase of 4.54%. The stresses in front of the excavation working face were coupled with the elliptical high-stress area in a secondary superposition. The stress increased from 16 MPa to 17 MPa, an increase of 6.25%. In summary, when the roadway is close to the center of the collapse column, the stress in the center of the column is low, and less damage occurs to the surrounding rock. There is a significant increase in stress on the sides and front of the roof. Vertical stress distribution of surrounding rock when exposing collapse column. (3) Analysis of stress evolution law of surrounding rock after exposing collapse column. Figure 9 demonstrates that when the roadway excavation is 190 m, the stress in the roof is coupled with the elliptical high-stress zone, which changes from 23 MPa to 24 MPa, an increase of 4.34%. The stress in front of the excavation working face increased from 17 MPa to 20 MPa, an increase of 17.6%. The stress on both sides of the collapse column was strengthened from 24 MPa to 26 MPa, an increase of 8.3%, when the roadway excavation was 220 m. At this time, the roadway entered the ordinary surrounding rock with the roof providing a certain degree of support. The stress of the roof is strengthened, which makes a secondary reinforcement for the stress on both sides of the collapse column. When the roadway excavation was 230 m, the stress of the roof did not change and remained at 26 MPa. The roadway leaves the influence zone of the collapse column, and the stress coupling phenomenon gradually dissipates. When the collapse column was exposed after 30 m ~ 50 m, the stress in front of the excavation working face was maintained at 17 MPa, which was reduced by 15%. In summary, when the roadway passes through the collapse column for 10 m, it remains within the influence zone of the collapse column. However, once the roadway passes 30 m ~ 50 m after the collapse column, it exits the influence range of the collapse column. Vertical stress distribution of surrounding rock after the collapse column is exposed. Analysis and discussion on the distribution pattern of plastic zone in surrounding rock When the roadway excavation is 10 ~ 30 m, shear failure predominantly occurs in the surrounding rock, as shown in Fig. 10. Circumferentially distributed shear failure was produced around the ellipse of the collapse column, and extensive shear failure occurred at the top of the collapse column. When the roadway excavation reaches a depth of 50 m, the failure range of the roof extends to 2.4 m, while the failure range of the floor reaches 2.01 m, and the failure range of the right side is 1.9 m. A small amount of shear-tension failure occurs at the top and periphery of the collapse column. The closer the roadway is to the trapped column, the greater the impact of mining stresses, leading to shear-tension failure at both the upper and periphery of the collapse column. when the collapse column is exposed, the mining-induced stress and the internal loosening lead to much shear failure and a minor tensile failure at the top of the collapse column. Failure law of surrounding rock when collapse column is not exposed. Figure 11 illustrates the failure distribution to the surrounding rock when the collapse column is revealed. When the collapse column was exposed for 10 m, a large amount of shear failure occurred at the periphery of the collapse column. The damage range of the roof reaches 6.96 m, and the failure of the floor reaches 7.32 m. A large amount of failure occurs at the interface between the collapse column and the coal seam. When the collapse column is exposed for 70 m, the failure range of the roof reaches 4.57 m, which is reduced by 34.3%, and the failure range of the floor reaches 4.3 m, which is reduced by 41.2%. In summary, the tectonic zone at the periphery of the collapse column leads to stress concentrations. A large amount of failure occurred in the peripheral tectonic zone of the failure column under the disturbance of mining stresses. When the roadway passes a collapse column, tensile failure occurs in the shallow surrounding rock, while shear failure takes in the deeper surrounding rock. Overrun pre-grouting is required to enhance the strength of the surrounding rock. Failure law of surrounding rock when exposing collapse column. Figure 12 illustrates the failure law of surrounding rock after the collapse column is exposed. After the roadway traversed the collapse column for 10 m, a large amount of shear failure occurred at the junction between the collapse column and the coal seam, as shown in Fig. 12(a). The roadway is currently affected by the collapsed column. After the roadway passes through the collapse column for 30 m~ 0 m, there is a decrease in the failure range of the surrounding rock, with shear failure becoming predominant. The failure range of the roof reaches 1.79 m, and the failure range of the floor reaches 1.9 m, as shown in Fig. 12(b) and (c). The current state of failure in the surrounding rock remains stable and is not progressing further in depth. In summary, The closer the roadway is to the collapse column boundary, the more significant the impact on the collapse column tectonic zone. Shear failure mainly occurs in the surrounding rock prior to the exposure of collapse column, and tensile failure mainly occurs in the surrounding rock during the exposure of the collapse column. Within the collapse column and within a 10 m range outside of it, the plastic zone of the surrounding rock experiences shear-tensile failure. Enhanced support is recommended for this area of the surrounding rock. Failure law of surrounding rock after the collapse column is exposed. Control technology for surrounding rock during roadway passing collapse column Surrounding rock zoning support technology When the collapse column is not exposed, the closer the roadway is to the center of the trapped column, the higher the stress is in front of excavation face. The stress in front of the excavation face is coupled with the peripheral ring stress of the collapse column. When the 480material roadway is 10 m away from the collapse column, the roadway is in the high stress concentration area. The exposure of the collapse column in the roadway leads to a reduction in stress in front of the excavation working face, resulting in a significance occurrence of tensile failures at the interface between the collapse column and the coal seam. At this stage, the roadway exhibits instability in its plastic zone. After the 480material roadway passes the collapse column, stress coupling still occurs between the excavation working face and the peripheral ring stress of the collapse column within a 10 m range on both sides of the collapse column. Currently, the roadway is in the high stress concentration area, susceptible to failure. For this reason, in order to overcome the high stress concentration and plastic zone instability, this paper proposes the zoned support technology for the perimeter rock of the roadway over the collapse column, as shown in Fig. 13. Surrounding rock partition support technology. (1) Overcoming areas of high stress concentration in the roadway. The 10 m range on both sides of the collapse column is considered as the high stress influence area, where the integrated support technology of “pre-grouting + bolt cable mesh + steel shed” is employed. Bolt specifications are Φ22-M24-2400 mm left-handed threaded steel bolt. The cable specifications respectively Φ22-6300 mm cable. The influence area of the collapse column is spans from 60 m to 190 m, while the outer range beyond the influence area of the collapse column extends from 0 to 50 m and from 190 m to 240 m. Figure 14(a)-(b) illustrates the design of the support outside the zone of influence, with bolt spacing of 1 m, row spacing of 1 m, and a cable arrangement of “3-2-3”. Figure 14(c)-(d) illustrates the support design for the area affected by the collapse column, with bolt spacing of 0.9 m, row spacing of 0.9 m, cable arrangement of “3-2-3”, and the spacing of the steel shed is 0.9 m. The specification of industrial steel shed is 4600 × 3600 × 900 (beam×leg×shed distance) with the fork angle of the shed leg set at 80°. Concrete is poured in advance under the legs of the shed as a ground beam to prevent sagging of the legs and improve the stability of the industrial steel shed. The whole section is laid with a layer of steel mesh and woven fabric, while 15 pieces of steel mesh are specifically laid on the top beam for added reinforcement.Furthermore, the two gangs of sheds are strung tightly behind the leg walls with dowels and brakes. 480 Material roadway support design diagram. (2) Improve the stability of the roadway surrounding rock. When the collapse column is exposed, a large amount of tensile-shear damage occurs at the junction between the collapse column and the coal seam, and the bolt and cable are ineffective in providing anchorage in the broken surrounding rock. The overrun pre-grouting is carried out for the surrounding rock in the area, and the slurry infiltrates into the broken surrounding rock to improve its strength. The grouting pipes with a length of 5.5 m were pre-bedded in the roof and sides of the 480 Material roadway. The slurry was formulated using 425# silicate cement in combination with the cement grouting additive XPM, and the water/cement ratio of the cement slurry was 0.8:1. The arrangement of grout holes in the roof is “3-2-3”, and the angle, spacing, and depth of grout holes in the roof are shown in Fig. 15(a). The angle, spacing, and depth of the grout holes on the two sides are shown in Fig. 15(b). Design of grouting parameters for 480 material roadway. Displacement and failure pattern of the surrounding rock with different support schemes In order to verify the reasonableness of the perimeter rock zoning control technology, in this paper, the FLAC3D software structure structural unit cable was used to simulate the bolt and cable support. Geogrid structural units were employed to simulate the flexible support of mesh. Beam girder units were used to simulate industrial steel shed support. The effect of grouting was simulated by altering the strength of the surrounding rock. The support parameters cover a wide range as documented in literature42,43. The parameters used in this paper for numerical simulation of the supporting structural units are specified in Table 4. Table 4Material parameters of support structures.Structural unitElastic modulus/GPaDiameter/mmPoisson ratioCement paste stiffness/N·m-2Cement slurry cohesion/N·m-1Bolt(Cable)210226.7 × 1092.4 × 105Mesh200.25Steel shed25.50.3 Material parameters of support structures. Figure 16 demonstrates the deformation law of the surrounding rock under different support schemes. A represents “bolt cable mesh” integrated support technology, while B represents “pre-grouting + bolt cable mesh” integrated support technology, and C represents “pre-grouting + bolt cable mesh + steel shed” integrated support technology. Due to the symmetry of the model, 480 material roadway was analyzed in the range of 0 to 120 m. When the collapse column is not exposed, the surrounding rock under the three support schemes is controlled by stability. The roof displacements in the three support schemes are 82.4 mm, 72.7 mm, and 63 mm respectively. Considering the support cost, scheme A is the optimum, as shown in Fig. 16(a). Figure 16(b) demonstrates the roof displacements for the three support schemes in 291.03 mm, 255.33 mm, and 214.52 mm. The roof deformation was reduced by 26.2% in scheme C compared to scheme A. When the roadway is excavated to a depth of 60 m, it enters the high-stress concentration zone, where the surrounding rock stress is coupled with the peripheral stress of the collapse column. A large amount of tensile failure occurs at the junction between the collapse column and the coal seam, and scheme C is considered the most effective. Grouting improves the strength of the surrounding rock, while the bolt serves as anchors to secure the intact surrounding rock. Figure 16(c) demonstrates the roof displacements for the three schemes with values of 561 mm, 567 mm, and 464 mm respectively, indicating that scheme C exhibits the most effective. When the roadway was excavated for 120 m, the 480-material roadway was located inside the collapse column, and the surrounding rock was in the influence zone of the collapse column. The surrounding rock in the area of the collapse column is broken, and grouting alone cannot effectively enhance the strength of the surrounding rock to ensure its stability. Comparatively, schemes A and B exhibit similar supporting effects. Therefore, scheme C is adopted, combining grouting to improve the strength of the surrounding rock with the “bolt cable mesh” and industrial steel shed as synergistic support measures to control the stability of the surrounding rock. Deformation law of roof under different support schemes. Figure 17 demonstrates the evolution of the plastic zone before and after the support for different excavation distances. When the collapse column is not exposed, the 480-material roadway surrounding rock mainly undergoes shear failure, with the failure range of the roof reaching up to 2.4 m and the failure range of the floor reaching up to 2.01 m, as shown in Fig. 17(a). The " bolt cable mesh " integrated support is adopted at the distance of 50 m of the roadway, resulting in a reduction in the failure range of the roof. This approach effectively achieves anchoring effects by means of the bolts and cables, as shown in Fig. 17(b). When the roadway is excavated for 60 m, the surrounding rock is in the high-stress concentration area. The tensile failure occurs in the roof and at the interface between the coal seam and the collapse column. The failure ranges of the roof, and the two sides are 5.2 m and 5.8 m, respectively, as shown in Fig. 17(c). The grouting pipe with a length of 5500 mm was used to grout the roof and the two sides to improve the strength of the surrounding rock. Finally, the integrated support of " pre-grouting + bolt cable mesh + steel shed " was adopted to control the stability of the surrounding rock, as shown in Fig. 17(d). Figure 17(e) and (f) demonstrate the distribution of plastic zones before and after the surrounding rock support when the roadway excavation is 120 m. The surrounding rock lies within the influence zone of the collapsed column, which has a loose interior and low strength of the surrounding rock. The integrated support of " pre-grouting + bolt cable mesh + steel shed " was adopted to control the surrounding rock effectively. Distribution of plastic zone before and after support at different excavation distances. Field applications In order to validate the effectiveness of the above zonal support technologies in controlling the stability of surrounding rock in the 480-material roadway, three monitoring stations were arranged in the 480-material roadway. Monitoring station #1 is 20 m away from monitoring station #2, and monitoring station #3 is 60 m away from monitoring station #2, as shown in Fig. 18. Layout of monitoring stations in 480-material roadway. The data collected from the three monitoring stations were collated to establish the correlation between the displacement of the surrounding rock and the monitoring time, as shown in Fig. 19. Figure 19(a) illustrates the deformation of the two sides and the roof at monitoring station #1. The maximum deformations are 68 mm, 64 mm, and 106 mm, respectively. The deformation of the roof and the two sides tends to be stabilized in about 20d. It can be seen that the 480-material roadway in this area can effectively be controlled by using the " bolt cable mesh” integrated support. The maximal deformation of the roof and the two sides at the #2 monitoring station was 100 mm, 102 mm, and 205 mm, respectively, as shown in Fig. 19(b). The roof exhibited the most significant deformation, with an increase of 93.4% in this area compared to monitoring station #1. The effective control of the surrounding rock deformation by the integrated support of “pre-grouting + bolt cable mesh + steel shed” for the 480-material roadway in this location was demonstrated at approximately 24 d. Figure 19(c) illustrates the maximum deformations of the roof and the two sides in monitoring station #3, which were 115 mm, 112 mm, and 220 mm respectively. The roof exhibited the greatest deformation, with a 7.3% increase in deformation observed in this area compared to monitoring station #2. In about 22d, the surrounding rock has been controlled, which proves that the integrated support of " pre-grouting + bolt cable mesh + steel shed " can effectively control the surrounding rock deformation in this area. Curve of deformation and time variation of surrounding rock. Conclusions The present study focuses on the simulated fractured and loosely nonlinear features in the collapse column. This simulation utilizes the Discrete Fracture Network (DFN) modeling and a strain-softening constitutive model. The size of the DFN follows a complicated power-law distribution, and the location of fractures within the DFN follows a uniform distribution. Additionally, the orientation of the DFN follows a Fisher distribution. The parameter calibration establishes the softening function for the shear strain in the collapse column in relation to the angle of friction and cohesion, as well as the softening function for the tensile strain in relation to the tensile strength. The present study focuses on the simulated fractured and loosely nonlinear features in the collapse column. This simulation utilizes the Discrete Fracture Network (DFN) modeling and a strain-softening constitutive model. The size of the DFN follows a complicated power-law distribution, and the location of fractures within the DFN follows a uniform distribution. Additionally, the orientation of the DFN follows a Fisher distribution. The parameter calibration establishes the softening function for the shear strain in the collapse column in relation to the angle of friction and cohesion, as well as the softening function for the tensile strain in relation to the tensile strength. The stress zones of the surrounding rocks around the collapse column are the internal impact zone of the collapse column, the stress decreasing zone, the stress increasing zone, and the initial rock stress zone. When the collapse column is exposed, the surrounding rock undergoes tensile damage. The closer the roadway is to the collapse column, the higher the stress is in front of the excavation working face and roof. The stresses on both sides of the collapse column are enhanced, and the stress of the roof is coupled and superimposed with the high-stress zone of the collapse column. The stress zones of the surrounding rocks around the collapse column are the internal impact zone of the collapse column, the stress decreasing zone, the stress increasing zone, and the initial rock stress zone. When the collapse column is exposed, the surrounding rock undergoes tensile damage. The closer the roadway is to the collapse column, the higher the stress is in front of the excavation working face and roof. The stresses on both sides of the collapse column are enhanced, and the stress of the roof is coupled and superimposed with the high-stress zone of the collapse column. The impact zone of the collapse column is 60–190 m, while the area outside the influence of the collapse column is 0–50 m, 190–240 m. The zone support technology is proposed, and the " pre-grouting + bolt cable mesh + steel shed " integrated support is adopted for the impact zone of collapse column, and the " bolt cable mesh " integrated support is adopted for the roadway outside the influence zone of the collapse column. After on-site practice, it has been observed that within approximately 20 days, stabilization in terms of deformation occurs in the roof and the two sides of the 1–3# monitoring station. Furthermore, effective control over surrounding rock has been achieved with practical supporting effect. The impact zone of the collapse column is 60–190 m, while the area outside the influence of the collapse column is 0–50 m, 190–240 m. The zone support technology is proposed, and the " pre-grouting + bolt cable mesh + steel shed " integrated support is adopted for the impact zone of collapse column, and the " bolt cable mesh " integrated support is adopted for the roadway outside the influence zone of the collapse column. After on-site practice, it has been observed that within approximately 20 days, stabilization in terms of deformation occurs in the roof and the two sides of the 1–3# monitoring station. Furthermore, effective control over surrounding rock has been achieved with practical supporting effect. Acknowledgements The author(s) would like to thank all editors and anonymous reviewers for their comments and suggestions. This research was financially supported by the National Natural Science Foundation of China (Grant Nos. 52074146, 52004117 and 52174117) Liaoning Province Doctoral Youth Seedling Project (LJ212410147009). Author contributions Ce Jia: Data curation, Writing - original draft. Chaojun Fan and Sheng Li: Writing guidance, plan guidance. Xiaoshuai Liu: Chart drawing. Mingkun Luo: On site experimental design and guidance. Data availability The data that support the findings ofthis study are available from thecorresponding author upon reasonable request. The software FLAC3D 6.0 was utilized in this study and can be accessed through the following link: https://www.itascacg.com/software. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The data that support the findings ofthis study are available from thecorresponding author upon reasonable request. The software FLAC3D 6.0 was utilized in this study and can be accessed through the following link: https://www.itascacg.com/software.