Research on the Deviatoric Stress Mode and Control of the Surrounding Rock in Close-Distance Double-Thick Coal Seam Roadways

Abstract

To address the issue of sustained deformation in the main roadway surrounding rock triggered by intense movement of overlying strata following the reduction of width of the stopping pillar (WSP) in closely spaced double extra-thick coal seams (CSDECS). Analyze the evolution patterns of abutment pressure, principal stress vector lines, and zones of deviatoric stress concentration (ZDSC) of the main roadways using multi-method approaches. The findings are as follows: As the WSP is reduced, the maximum abutment pressure (MAP) on both sides of the gate roadways’ surrounding rock becomes significantly more asymmetric and intense. The deflection trajectory of the maximum principal stress line (MPSL) in the two coal seams, induced by the advancing underlying panel, follows an approximate inverted ︺ shape. The evolution of the ZDSC and the main roadways in the adjacent working faces all shows three-stage characteristics. For the upper coal seam, it is characterized by crescent-shaped symmetry → slow and asymmetric increase of the peak value and the offset of the ZDSC → the ZDSC on the non-mining side (NM-S) reaches the maximum while the mining side (M-S) shows the reverse trend. For the lower coal seam, it is characterized by crescent-shaped symmetry → quasi-annular distribution with a slight increase in the peak value → significant and asymmetric increase of the peak values. Based on the identification of the key control zones in the ZDSC, an asymmetric reinforcement segmented control method was proposed. The findings provide useful guidance for analogous engineering projects.

1. Introduction

Thick and extra-thick coal seams serve as the primary layers for high-throughput and energy-efficient. However, compared with thin and medium-thick seams, the intense overlying strata movement, resulting from the significantly increased mining height, poses a serious challenge for main roadway control [1]. In mining close-distance coal seams, it is necessary to solve complex issues, such as the combined damage to the main roadway caused by the superposition of the mining stress fields from the two coal seams. When different forms of two types of coal seams coexist, producing complex geological conditions with closely spaced, double extra-thick seams, mining must face the challenges posed by the occurrence conditions of each seam. Additionally, focus should be on stabilizing the roadway and establishing a proper WSP to ensure safe mining under these geological conditions.

While an excessively WSP left to stop mining greatly reduces the mining-induced impact on the roadway, it leads to significant coal resource losses. Conversely, an excessively narrow coal pillar for stopping mining recovers more coal resources but increases the mining impact on the main roadway, causing continuous deformation and significantly raising the difficulty of control. Only by clarifying how the surrounding rock stress redistributes and understanding the characteristics of biased overload modes on the main roadway—after being affected by mining under different conditions of the WSP—can we identify the key control area in the roadway. Strengthening control within this area allows for the scientific determination of the optimal WSP, leading to the efficient recovery of coal resources.

Scholars have conducted in-depth research on various aspects, such as the overlying rock structure and fracture mechanism under the mining conditions of single thick and extra-thick coal seams [2,3,4,5] the stress evolution law of mining roadways under strong mining disturbance [6]. the deformation and failure mechanisms [7,8]; roadway control [9,10,11]; the instability and control of main roadways [12]; and the determination of stopping coal pillar widths [13]. Additionally, significant research has been conducted in the following areas: the movement of overlying rock structures and the evolution of fractures under the mining conditions of closely spaced thin and medium-thick coal seam groups [14,15,16], the failure mechanism of floor rock strata caused by overlying coal pillars [17,18], the instability mechanism [19] and control [20,21,22] of mining roadways, the optimization of roadway layout [23], and the stress evolution [24], deformation, and control [25] of the main roadways. However, research on the deformation of main roadways and the WSP in conditions involving the CSDECS is still limited. Meanwhile, mining these tightly packed, large coal seams differs significantly from traditional thick-seam mining or groups of closely spaced, thin, or medium-thick coal seams. With the same WSP, the extent and scope of the resulting mine pressure behavior are more noticeable. Additionally, the main roadway is different from the mining roadway, and it has a longer service life. If substantial deformation occurs, it not only requires repeated side brushing, which is time-consuming and labor-intensive, but also seriously affects the smooth operation of the mine roadway system and could even halt the entire production process.

Currently, most research on the width setting of coal pillars for roadway stopping and the deformation control of surrounding rock in main roadways focuses on analyzing how the vertical stress of the surrounding rock evolves. Additionally, most surrounding rock control schemes are symmetrical. However, there is little research on the relationship between the deflection law of the principal stress vector, the distribution pattern of deviatoric stress, and the control of key areas in the surrounding rock of main roads. Currently, research on the WSP for main roadways and the deformation predominantly focuses on analyzing the evolution of vertical stress, with control measures often adopting symmetrical designs. However, under the conditions of CSDECS, studies remain scarce regarding the deflection of principal stress vectors in main roadways, the distribution patterns of ZDSC, and their relationship with key areas of surrounding rock control. Furthermore, the crucial support areas of the roadways are identified to enable scientific control and to optimize the WSP.

It is crucial to gain a deeper insight into the deviatoric stress evolution and its control in main roadways, which is influenced by varying WSP in CSDECS.

2. Engineering Background

The main mining areas are the 4^# coal seam and 6^# coal seam in the mine, both with dip angles between 2° and 5°, which are considered near-horizontal coal seams. According to the spatial relationship between mining and geological drilling, as shown in Figure 1. With average thicknesses of 11.0 m and 13.5 m, respectively, the 6^# coal seam is situated 21 m beneath the 4^# seam and has a burial depth of 300 m. Because there are buildings on the early surface that need to be protected by coal pillars, the WSP between Panel 24103 and the return air gateway in the 4^# coal seam has a width of 120 m. In recent years, with the removal of surface buildings, research has been conducted on whether the 6^# coal seam has the potential to optimize the WSP. After the extraction of the 26103 longwall panel, which was mined by the fully mechanized sub-level caving method (3.4 m cut and 10.1 m caving height), the main haulage roadway must be preserved to ensure it continues to serve production needs, as shown in Figure 1.

3. Simulation and Analysis of Stress Evolution of the Main Roadways in CSDECS

To investigate whether the 6^# coal seam can optimize the WSP and further enhance the recovery rate of coal resources during close-distance double extra-thick coal seam mining, a study was conducted. With the geological conditions, a numerical model with varying WSP was created. The evolution of abutment pressure, the deflection characteristics of the principal stress vector, and the partial load mode in the ZDSC were analyzed.

3.1. The Establishment of the Model

Based on the mining space relationship diagram and geological drilling map, the FLAC3D v7.0 numerical model is built. As shown in Figure 2, the average burial depth of the 6^# coal seam is 300 m. The model had a height of 86.5 m from the 6^# coal seam floor, with an additional 213.5 m of overburden not included in the simulation. Based on the formula Pz = γH, a vertical load of approximately 6.25 MPa is applied to the top boundary of the model to simulate the overlying strata. The horizontal stress is assigned according to the actual lateral pressure coefficient of 1.2. Boundary conditions are as follows: sides—normal constraint; bottom—fixed. Utilize a strain-softening model for the coal seam and the Mohr–Coulomb criterion for the rock strata. The return airway of the two coal seams is rectangular and excavated along the roof. The main and auxiliary transport roadways for each coal seam are all straight-walled, semi-circular arch roadways along the floor, and the return airway of the 4^# coal seam is closest to the mining face.

3.2. Evolution of the Abutment Pressure Field of the Main Roadway in CSDECS

This paper addresses the challenge of the CSDECS. Through numerical modeling, which can analyze the effect of WSP (120 m to 20 m) on main roadway stability. Key insights include the evolution and distribution of bearing pressure in the roadways of both upper and lower seams during longwall retreat, with details depicted in Figure 3. As illustrated in Figure 3, as the working face advanced from 120 m to 20 m from the 4^# coal seam return airway, a non-symmetric increase in the MAP on both sides of the upper and lower seam gate roads was observed due to intensified mining-induced effects.

(1)

As the WSP decreases, the outward shift of the peak abutment pressure in the return air roadways of the CSDECS becomes much higher than that in the surrounding rock of other roadways. The range of the fractured zone in the return air roadways of the two coal seams expands significantly. This indicates that, during the process of reducing the WSP, controlling the surrounding rock of the return airways of the CSDECS is the key research focus.

(2)

Since the 4^# main roadway, 6^# main roadway, and 6^# auxiliary roadway are far from the 26103 panel, the MAP on the sides of these main roadways does not change significantly due to mining throughout the entire process. When the WSP is less than 40 m, the MAP on the sides of the 4^# auxiliary roadway shows an asymmetric increase. Additionally, the stress increase on the M-S of the roadway (denoted as the R-side, and similarly hereafter) is about 1.8 MPa higher than on the NM-S (denoted as the L-side). Comparing the changes reveals that, since the 4^# return airway is closest to the working face, it is significantly affected by mining activities. The MAP on the sides of the 4^# return airway shows a marked asymmetric increase when the WSP is less than 80 m. The MAP on the R-side of the 4^# return airway remains higher than on the L-side. While reducing the WSP, the MAP increases by approximately 33.5% and 38.1%, respectively, due to the effects of superimposed mining. Compared to the 4^# return airway, as the WSP is less than 100 m, the MAP on sides of the 6^# return airway also shows an asymmetric increase due to mining, with a higher stress increase than that observed on the 4^# return airway.

3.3. Evolution of Principal Stress Vector Deflection in the CSDECS

The above analysis has examined the evolution characteristics and load-increasing laws of abutment pressure fields of the upper and lower coal seam roadways, considering different WSP when stopping mining in the roadways of CSDECS, which offers guidance for engineering practice. In fact, nine stress components—or alternatively, three principal stresses—define the stress state at any point. The abutment pressure only reflects a single vertical stress, which restricts its usefulness in analyzing complex problems. Therefore, the following analysis will be performed according to the principal stress field. Which not only reveals the principal stress field behavior in the upper and lower coal seams when the underlying working face of the CSDECS approaches the roadway continuously but also allows for the analysis of the mechanical response through the deflection effect of the max principal stress vector (MPSV). Therefore, the mathematical definition of the MPSV angle (α) is specified: it is defined as the angle between the MPSV and the vertical line (range: −90° ≤ α ≤ 90°), as shown in Figure 4.

In Figure 5, as the underlying advanced working face continually approaches the main roadway, the deflection of the MPSV of the upper and lower coal seams is as follows:

(1)

As the WSP decreases, the deflection patterns of the MPSV of the upper and lower coal seams are as follows: towards the solid coal side → nearly horizontal direction → towards the goaf side. The approximate trace of the spin of the MPSV resembles the shape of “︺”.

(2)

The angle α gradually increases, ranging from 25° to 65°, with a difference of 40°. The angle α in the lower coal seam first increases and then decreases, with a change magnitude greater than that of the upper seam. The lower coal seam exhibits greater variability, the physical essence of which stems from the multiple superposition, transfer, and redistribution processes of mining-induced stress. Specifically, the extraction of the upper seam causes the initial stress field to undergo its first redistribution, resulting in a pattern of low pressure in the goaf area and high pressure in the coal pillar zones. Simultaneously, the stress concentration effect induced by the coal pillars in the upper seam transfers downward to the lower coal seam. Although the lower seam has not yet been mined, its original in situ stress state has already been altered due to the influence of upper seam mining, disrupting the initial equilibrium. When mining advances to the lower seam, the stress field undergoes a second redistribution, particularly in areas affected by the remnant coal pillars from the upper seam. The longwall face in the lower seam is subjected to two superimposed stresses: one is the high stress transferred downward from the coal pillars of the upper seam, and the other is the front abutment pressure generated by the extraction of the lower seam itself. This dual stress superposition effect results in a stress environment significantly different from that of a single ultra-thick coal seam, thereby leading to stronger variability in the lower seam mining conditions.

3.4. The Evolution of ZDSC in CSDECS

The deviatoric stress differs from the abutment pressure. It not only involves the three principal stresses but also, according to elastoplastic theory, governs the plastic deformation after shear. Therefore, introducing the deviatoric stress as an analysis index can comprehensively reflect the trend of the plastic development of the main roadway adjacent to the 26103 panel, being affected by mining and whose influence intensifies during the process of reducing the WSP in the large-section roadway of CSDECS based on elasticity [26]. This paper examines the max principal deviatoric stress, which plays a pivotal role in the stress tensor, as shown in Equation (1).

$$\begin{array}{l}{S}_1={\sigma }_1-{\displaystyle \frac{{\sigma }_1+{\sigma }_2+{\sigma }_3}{3}}\\ S_3={\sigma }_3-{\displaystyle \frac{{\sigma }_1+{\sigma }_2+{\sigma }_3}{3}}\end{array}$$

(1)

$$\begin{array}{l}{\sigma }_1={\displaystyle \frac{{\sigma }_r+{\sigma }_{\theta }}{2}}+{\displaystyle \frac{1}{2}}\sqrt{{({\sigma }_r-{\sigma }_{\theta })}^2+4{\tau }_{r\theta }^2}\\ {\sigma }_3={\displaystyle \frac{{\sigma }_r+{\sigma }_{\theta }}{2}}-{\displaystyle \frac{1}{2}}\sqrt{{({\sigma }_r-{\sigma }_{\theta })}^2+4{\tau }_{r\theta }^2}\end{array}$$

(2)

where h is the depth; λ is the lateral pressure coefficient; a is the tunnel radius; and the polar coordinates of any point are set as (θ, r). The above equation presents a simplified deviatoric stress formula for plane problems in elastic mechanics [26].

The calculation formula for the radius of the plastic zone of a circular roadway in a non-hydrostatic stress field is as follows [26]:

$$r=\sqrt{{\displaystyle \frac{2a^2\gamma h(1-\lambda )\sin \phi \cos 2\theta }{2C\cos \phi +\gamma h(1+\lambda )\sin \phi -(s_1-s_3)}}}$$

(3)

The correlation between deviatoric stress and roadway failure can be quantified by substituting the principal deviatoric stress S₁ into the boundary of the plastic zone of the circular roadway in the non-hydrostatic stress field.

During the WSP (120 m to 20 m) from the airway of 4^# coal seams, after the WSP is reduced, the ZDSC of the return airways in the upper and lower coal seams are analyzed.

From Figure 6 and Figure 7, the laws of the ZDSC in the return airways in the coal seams under different WSP are as follows:

(1)

When the WSP (120 m to 20 m) is away from the return airway of the 4^# coal seam, the deviatoric stress max values (DSMV) curves on sides of the airways exhibit significant asymmetric loading increase. The amplitude of the increase in the DSMV of the 4^# coal seam’s return airway is notably higher than that of the 6^# coal seam’s return airway, indicating that the shear failure degree of the 4^# airway is much greater than the 6^# airway.

(2)

The distribution of the DSMV of the return airway in the 4^# coal seam exhibits a spatial dynamic evolution law with increasing mining disturbance. As the WSP is reduced from 120 m to 50 m, the ZDSC of the roadway surrounding rock rotates counter-clockwise. Its distribution is specifically as follows: “Both sides of the return airway → Transition area from the bottom corner on the L-side to the shoulder corner on the R-side → Transition area of the roof—Transition area of the bottom corner on the R-side → Transition area of the shoulder corner on the L-side—Transition area of the bottom corner on the R-side,” indicating that after the WSP is decreased, the plastic development trend does not exhibit a symmetrical form. Therefore, the traditional symmetrical support structure of the roadway cannot effectively control the deformation. Meanwhile, when the WSP is 40 m, the ZDSC undergoes a sudden change. The ZDSC of the R-side is connected with the stope, resulting in a sharp increase in peak deviatoric stress intensity and a significant expansion of the peak area range. When the WSP is reduced to 20 m, the ZDSC of the roadway connects with the stope. At this point, the mining influence has shifted to the L-side, while the R-side suffers damage from strong shear forces, causing the DSMV on the L-side to be higher than the other. As the WSP decreases from 50 m to 20 m, the ZDSC on the L-side does not shift, but its range and DSMV increase significantly. Conversely, the ZDSC of the R-side decreases, and the plastic undergoes significant expansion. Eventually, the R-side transforms from a plastic zone to a low-strength bearing zone, and the load-bearing capacity drops considerably. Ultimately, this leads to continuous asymmetric deformation of the roadway, causing support failure.

(3)

The distribution and patterns of the ZDSC of the return airway in the 6^# coal seam differ significantly from the 4^# coal seam. Specifically, when the WSP decreases from 120 m to 20 m, the ZDSC of the 6^# airway rotates counterclockwise. Meanwhile, the ZDSC is approximately crescent-shaped. Its evolution law is “the transition area from the bottom corner on the L-side to the shoulder corner on the R-side → the roof–floor → the transition area from the shoulder corner on the L-side to the bottom corner on the R-side → the left and right sidewalls”. Since there is still a 20 m WSP of return airway in the 6^# coal seam and the 4^# coal seam, as the WSP decreases, the behavior of the ZDSC of the 6^# return airway does not respond as significantly as that of the 4^# return airway. As mining of the section progresses, the crescent-shaped ZDSC of the airway remains unchanged, indicating that the stability of the 6^# airway is better than that of the 4^# airway.

In summary, when the WSP exceeds 50 m, the 4^# airway plastic zone is not connected to the stope. However, if the WSP is less than 50 m, the 4# airway plastic zone is connected to the stope. Currently, bolt and cable support are insufficient to maintain the roadway’s stability. Because the WSP between the 26103 panel and the 4^# return airway has been optimized from 80 m to 50 m. Additionally, only reinforcement support for the return airway of the 4^# coal seam is required.

4. Similarity Test of Closely Spaced Double Extra-Thick Coal Seams

To further investigate the behavior of deformation and failure of the adjacent main roadways under different WSP in CSDECS, a similar test was conducted to simulate the effects of mining activities on the 4^# return airway from the 24103 and 26103 working faces. According to the size of the experimental area and the test bench, the geometric similarity constant ratio of this model is 1:100, and the unit weight (1:1.14) and stress (1:150)—which are not numerically identical but were tailored through balancing model material properties with prototype response to achieve equivalent strength–stiffness scaling. During testing, material variation was monitored (density deviations <±2%) under standard curing conditions (20 ± 2 °C, >95% humidity) to ensure consistency. Graded loading was applied at a scaled rate (time ratio 1:10), maintaining synchronization in a mining engineering context.

In the test, sand, lime, and gypsum are used as the main proportioning materials, and mica sheets are used as the layering materials. The names and proportioning conditions of each layer of the model are shown in

Table 1. Ratio data of layered coal and rock materials.

	Stratum	Thickness/mm	Ratio	Sand/kg	Lime/kg	Gypsum/kg	Water/kg
1	Immediate floor	100	6:0.5:0.5	13.1	1.1	1.1	1.53
2	6# Coal	130	8:0.7:0.3	20.5	1.8	0.8	1.96
3	Medium Sandstone	70	7:0.6:0.4	9.4	0.79	0.79	1.17
4	Siltstone	80	6:0.5:0.5	10.8	0.9	0.9	1.28
5	Siltstone (Sandstone)	50	6:0.5:0.5	6.7	0.56	0.56	1.07
6	4# Coal	110	8:0.7:0.3	17.3	1.5	0.7	1.68
7	Mudstone	30	6:0.5:0.5	4.3	0.36	0.36	0.9
8	Sandy Mudstone	30	6:0.5:0.5	4.3	0.36	0.36	1.03

.

With the WSP of 80 m to 20 m from the 4^# return airway, the deformation of the airway under three conditions with coal-pillar widths of 60 m, 40 m, and 20 m in the airway’s coal-stopping area was analyzed, as shown in Figure 8.

According to the results of similar simulation tests, when the WSP was 60 m from the 4^# return airway, the deformation of the airway’s roof and the bottom corner on the mining side was relatively large; when the WSP was 40 m from the 4^# return airway, the deformation of the non-mining side roof transition area and the mining side floor transition area of the airway was relatively large. With the 20 m WSP of the 4^# return airway, the deformation range of the NM-S roof transition area and the M-S floor transition area of the airway further increased. Meanwhile, the analysis showed that the deformation area of the airway was consistent with the range of the ZDSC in the numerical simulation.

5. Evolution Mode of Partial Additional Load in the ZDSC in the Main Roadway of CSDECS

From the above numerical simulations and similar tests of the main roadway in CSDECS under different WSP, it can be seen that as the WSP decreases, the plastic development around the main roadway becomes asymmetric. The traditional symmetric support method cannot effectively support the key control area of the main roadway. Therefore, studying the mode of the partial additional load in the ZDSC in CSDECS is essential. It can help predict the plastic development of the main roadway, enhance control over its range, and has significant practical value for guiding engineering applications.

As shown in Figure 9, when the WSP decreases from 120 m to 20 m, the influence of the 26103 panel mining on the return air roads of the coal seams becomes more intense. Consequently, the eccentric load mode of the ZDSC is as follows:

(1)

When the WSP is relatively large, the disturbance to the return air main roads of the two coal seams is relatively small. The ZDSC of the main roads exhibits a crescent-shaped, centrosymmetric distribution. The ZDSC of the return airway for the 4^# coal seam is mainly distributed. In contrast, the ZDSC of the return air main road for the 6^# coal seam is located in the transition zone from the bottom corner on the L-side to the top corner on the R-side. Additionally, the range of the ZDSC in the 6^# return airway is larger than that of the 4^# return airway. However, since the return airways of the two coal seams are both far from the working face and there is a sufficiently wide protective coal pillar to isolate the strong dynamic load impact caused by mining, the mining disturbance they experience is relatively small, and the conventional support can meet the requirements for the deformation stability of the roadways.

(2)

As the WSP decreases from a medium size, the level of mining disturbance on the return air roadways of the two coal seams gradually increases. The ZDSC of the roadways shifts from a centrally symmetric crescent shape to an asymmetric, expanding shape. Similarly, because of the spatial position difference between the two roadways, the levels of increased disturbance vary, and there are differences in the loading patterns of the ZDSC. Specifically, the ZDSC of the 4^# return air roadway shifts from both sides toward the transition zone in the right-hand roof corner. While the ZDSC of the 6^# return air roadway deviates from the “transition area between the L-side bottom corner and the R-side top corner” to the “roof and floor,” forming a “quasi-annular high-deviatoric-stress area.” The counterclockwise deflection angles of the ZDSC of the 4^# return air roadway differ from those in the 6^# coal seam, indicating that the plastic zones develop along divergent paths after mining. Therefore, the key support areas should concentrate on the distribution range of the ZDSC.

(3)

As the WSP is relatively small, further reducing the WSP intensifies the mining disturbance on the return air roadways of the coal seams. The differences in the evolution of the ZDSC are significant. The ZDMV of the L-side in 4^# the return air roadway reaches its max, while the ZDSC on the R-side shows a reversal. Furthermore, these peak areas are primarily distributed in the transition zones between the roadway side and the roof, as well as between the roadway side and the floor, with the peak areas moving outward and increasing simultaneously. The reversal of the deviatoric stress peak area on the R-side indicates that this side is connected with the stope. At the same time, both the peak value and the range of the ZDSC of the L-side increase, suggesting that the mining influence has shifted from the R-side to the L-side of the roadway. Conversely, the deviatoric stress peak areas around the 6^# return air roadway exhibit significant asymmetric growth, and the shape of the peak area does not reverse, indicating that the 4^# return air roadway is more severely affected by mining.

(4)

The disparity arises from the mechanical effect that prior extraction of the 4^# coal seam induces stress release in the goaf and stress concentration in coal pillars, which transfers downward and causes heterogeneous stress disturbances in the 6^# coal seam. During the mining of the 6^# coal seam, as the 26103 panel advances beneath these pillars, the concentrated stress superimposes with mining-induced stress, resulting in a significant increase in the roadway. The difference in stress distribution primarily stems from the mining sequence and spatial positioning: stress transfer in th 4^# coal seam is dominated by downward transmission from coal pillars, while the 6^# coal seam experiences a more complex and intense perturbation due to the superposition of its own mining stress and residual stress from the upper layer.

6. Engineering Tests and Mine Pressure Monitoring

Based on the above research, as the WSP is reduced to 50 m, the position differences of the main roadways cause the 4^# return airway to be closer to the mining face compared to the 6^# return airway. The ZDSC of the 4^# coal seam return airway due to mining shows a significantly higher level of intensification. For the 6^# return airway, with a 75 m WSP, the surrounding rock fracture zone is not fully developed, and the initial scheme can maintain its stability. For the 4^# return airway, the surrounding rock is more heavily fractured. Meanwhile, with the WSP at 40 m, the ZDSC connects with the stope, making support difficult. Therefore, the 4^# return airway is chosen as the test roadway, as shown in Figure 10.

6.1. Support Strategy and Scheme for the 4^# Return Airway

The plasticization of the rock mass occurs due to deviatoric stress. The surrounding rock between the ZDSC and the main roadway has entered the plastic stage, while the rock outside this peak area remains in the elastic stage. Therefore, supporting the 4^# main roadway requires penetrating through the peak area of the deviatoric stress, anchoring long cable bolts into the elastic area, and working in conjunction with the original support to squeeze, reinforce, and control the fractured surrounding rock. Meanwhile, the long cable bolts of the reinforcement support, along with the original anchor bolts and cables, strengthen the broken rock to improve its capacity. Additionally, the cable bolts can improve the shear failure resistance of the roadway. As the WSP of the 4^# return airway is 50 m, the ZDSC is concentrated in the roof on the L-side and in the transition zone between the R-side and the floor. The peak range on the L-side is larger than the M-S. Therefore, the support needs to perform key-area partitioned asymmetric support control for the ZDSC, as shown in Figure 10.

Original support: For the roof bolts and side bolts of the 4^# return airway, left-hand non-ribbed deformed steel bars with a Φ 20 × 2200 mm are used, with a spacing of 900 × 900 mm, and are equipped with square pallets of 125 × 125 × 10 mm. The roof cable bolts are made of steel strands with a Φ 18.9 × 6000 mm, equipped with square pallets of 250 × 250 × 16 mm, with a spacing of 1800 × 2700 mm.

Key-area partitioned asymmetric support scheme: Building on the original support plan, at a point 1 m from the roof on the L-side, an additional steel strand is installed at a 15° angle with the horizontal toward the roof. At a point 250 mm from the left side of the roof, an extra cable bolt is added at a 15° angle with the horizontal from the working face. On the R-S of the roadway, 1 m above the floor, an additional cable bolt is installed at a 15° downward angle toward the roadway floor. Simultaneously, at a point 500 mm from the right side of the floor, another cable bolt is installed at a 45° angle toward the M-S. An additional cable bolt is also installed right in the dead center. The diameter of these additional cable bolts is Φ 21.6 × 6600 mm, equipped with square plates measuring 300 × 300 × 16 mm, with a spacing of 1800 mm.

6.2. Mine Pressure Monitoring

To verify the control effect of partitioned asymmetric support in the key area of the 4^# return airway, measuring stations were established in the airway. During the process with the WSP 120 m to 50 m from the 4^# return airway and within 45 days after the mining was completed, on-site observations were conducted on roof subsidence, floor heave, and the convergence of the two side walls of the airway. In Figure 10, as the WSP is large, the overall deformation was minimal. As the working face continued to mine and moved closer to the airway, the disturbance intensity on the airway increased. The convergence of the 4^# return airway showed an increasing trend following an asymptotic exponential function. As illustrated in Figure 10, the on-site monitoring was conducted using a leveling staff (with a measurement accuracy of 1 mm). Convergence of the roof-to-floor and rib-to-rib was measured on a daily basis. The stations were spaced at 10 m intervals, with 32 measuring points uniformly arranged at each station. The total monitoring length covered 180 m. After the working face stopped mining, the deformation of the airway tended to stabilize. The roof subsidence stabilized at 29.82 mm, the floor heave at 48.91 mm, and the convergence of the two side walls at 98.15 mm. The field monitoring results indicate that, without implementing reinforcement support, the convergence in the key control zone reached 0.9–1.2 m, and the deformation trend of the surrounding rock in this area was generally consistent with the numerical simulation results. After reinforcement support was applied, the convergence deformation of the roadway was effectively controlled, with the convergence in the key control zone of the main roadway surrounding rock reduced by approximately 40% to 55%. These quantitative results fully verify the reliability of the model and provide empirical support for the effectiveness of the asymmetric support scheme.

7. Discussion

(1)

Different research indicators are selected, each representing distinct physical implications (as shown in

Table 2. Selection of stress values.

Stress Value Scale	Intrinsic Nature	Physical Significance	Application	Evaluation
Individual stress component (σxx, σyy, σzz, τxz, τxy, τyz, τzx, τzy, and τyx)	Components of the stress tensor	Stress in a specific direction within a given coordinate system	Original data describing the stress state	It does not fully characterize the stress state at a point
Single principal stress (σ1, σ2, and σ3)	Eigenvalues of the stress tensor	Extreme values of the stress state at a point	Applied to strength theories based on maximum tensile/compressive and shear stresses.	It does not fully characterize the stress state at a point
Principal deviatoric stress S1 $S_1={\sigma }_1-{\displaystyle \frac{{\sigma }_1+{\sigma }_2+{\sigma }_3}{3}}$	Eigenvalues of the deviatoric stress tensor	Extreme values of the stress component causing shape distortion	Fundamental for calculating J2 in plasticity analysis.	It can characterize the stress state at a point

). After comprehensive consideration, deviatoric stress is adopted as the primary research indicator. This is because it can not only characterize the stress state at a point but also represent the shear failure in the deep rock mass.

(2)

The research on the influence of different WSP in the underlying mining face of CSDECS on the stability of the main roadway is still lacking. Meanwhile, at present, most of the research on the setting of the width of coal pillars left during stopping mining in roadways and the control deformation is limited to the analysis of the evolution of the vertical stress of the roadway surrounding rock, and the surrounding rock control schemes are mostly symmetrical.

(3)

The traditional symmetrical support system for roadways lacks an understanding of “the influence of the deflection of the principal stress and the distribution of the ZDSC on the key control area of the roadway“. Therefore, through the research on the influence of different WSP left during stopping mining in the underlying face of CSDECS on the stability of the roadway, this paper obtains the deflection law of the principal stress vector and the deviatoric stress peak area’s incremental loading mode, determines the key control direction of the roadway, and proposes a partitioned asymmetric support method for roadways based on the deviatoric stress incremental loading mode. This method makes up for the limitation of using the traditional symmetrical support method for roadways regardless of the size of the coal pillars left during stopping mining and is of great significance for guiding engineering practice.

8. Conclusions

(1)

During the advancing process of the lower seam face in CSDECS, the orientation of the maximum principal stress in both the upper and lower seams deviates, following an approximate “︺”-shaped trajectory: the maximum principal stress vector line first “\”→ then “—” → last “/”.

(2)

The evolution of the ZDSC and the main roadways in the adjacent working faces all shows three-stage characteristics. For the 4^# coal seam, it is characterized by crescent-shaped symmetry → slow and asymmetric increase of the peak value and the offset of the ZDSC → the ZDSC on the NM-S reaches the maximum while the M-S shows the reverse trend. For the 6^# coal seam, it is characterized by crescent-shaped symmetry → quasi-annular distribution with a slight increase in the peak value → significant and asymmetric increase of the peak values.

(3)

By studying the ZDSC loading patterns under different WSP, key control areas were identified. Reinforcement support was applied at corresponding locations, ensuring gateway stability while optimizing the WSP to 50 m (a reduction of 30 m), significantly improving the resource recovery rate.

(4)

Based on the loading patterns of ZDSC corresponding to different WSP, an asymmetric support method for different zones was proposed. The approach breaks through the traditional symmetric support form and provides important guidance for engineering practice.