In the unmined disturbed area of the west three areas protruding appraisal of the No. 5 roadway 18 coal top and bottom plate each take 10 m rock core, study the roadway top and bottom plate of the location of different rock layers and layering thickness, access to the roadway perimeter rock structural combination of characteristics, so as to carry out the physical and mechanical properties of the rock of each rock layer test, the top and bottom plate rock core is shown in ( Figure 1 ).

From ( Figure 1(a) ), it can be seen that the 18-coal roof slab core is light gray, medium cementation, medium sorting, dominated by quartz and rock fragments in the range of 0 - 2.43 m drilling core, judged as medium-grained sandstone, and the integrity of this section of the core is poor in the range of 0 - 1.3 m due to the periphery of the roadway. There are two sections larger than 20 cm in the whole section: 4.0 - 4.54 m and 5.0 - 5.34 m. These two sections are also judged to be medium-grained sandstones. These two sections are broken, and only one section of the core is larger than 10 cm. Within the range of core 2.43 - 4.0 m, the rock is mainly grayish white, with medium cementation and poor sorting, and the composition is dominated by quartz, feldspar, and metamorphic rocks, which is judged to be a conglomerate-bearing coarse-grained sandstone, and within the range of the whole section, there are four sections of greater than 10 cm, with the cores being more broken, and the rest of the cores being more complete and more of them are short columns, 4.54 - 5.0 m, 5.34 - 6.0 m, 7.0 - 7.0 m, 7.0 - 7.0 m, and 7.0 - 7.0 m. 4.54 - 5.0 m, 5.34 - 6.0 m, 7.0 - 7.71 m, 8.0 - 8.86 m are also judged as coarse-grained sandstone. The whole is relatively broken, and there are 6 sections with cores larger than 10 cm; the rocks in core 6.0 - 7.0 m are mainly gray, moderately cemented, and contain charcoal compounds. This is judged to be siltstone with good integrity. The last 0.4 m is broken overall, brownish black, massive, containing a lot of charcoal compounds, and is judged to be mudstone; core 9.59 - 10.0 m is dark gray, massive, containing charcoal debris, and is judged to be silty mudstone. The core is dark gray, blocky, with carbonaceous debris, and judged to be siltstone. The whole is relatively broken, and there is no core over 10 cm.

From ( Figure 1(b) ), it can be seen that the first 4 m of the core is overall more broken, and there are only 4 segments larger than 10 cm. Within the range of 0 - 0.22 m, 1.54 - 2.0 m, 7.17 - 8.0 m of the core of the 18-coal bottom plate, the rocks are mainly grayish-white, with medium cementation and poor sorting, and the constituents are dominated by quartz, feldspar, and metamorphic rocks, and are judged to be conglomerate - bearing coarse - grained sandstones, and there are 2 segments larger than 10 cm in the whole range of the section. The cores 0.59 - 1.54 m, 2.0 - 4.0 m, 6.0 - 7.17 m, 8.0 - 9.19 m are light gray, moderately cemented, moderately sorted, and dominated by quartz and rock fragments, which are judged to be medium-grained sandstones, and the cores, due to the fact that they are in the roadway, are larger than 30 cm in the entire range of 1 section, and there are only 3 sections of the cores that are larger than 10 cm. Within the range of 0.22 - 0.59 m and 9.19 - 10.0 m, the core is dark gray, blocky, containing charcoal debris, and is judged to be silty mudstone. The whole is relatively broken, and there are no cores of more than 10 cm.

Two 8m top plate drill holes were drilled in No.3 lane of West Zone 3 for peeping. No.1 and No.2 drill holes were peeped as shown in ( Figure 2 ).

From ( Figure 2(a) ), it can be seen that the rock structure is more complicated, which are medium-grained sandstone, coarse-grained sandstone and siltstone. The stability of the roof is good, the transverse fissures mainly appear in the range of 0 - 3 m, which are prone to delamination, and the transverse fissures can be seen at the positions of 0.5 m, 1.53 m, 1.65 m, 1.92 m, 2.06 m, and 2.72 m. The rest of the positions are more intact, and the fissures are moderately developed.

From ( Figure 2(b) ), it can be seen that the rock structure is more complicated, which are medium-grained sandstone, coarse-grained sandstone and siltstone mudstone. The stability of the roof is good, the transverse fissures mainly appear in the range of 0 - 3 m, which is easy to occur off the layer, the transverse fissures can be seen at the position of 0.05 m, 0.33 m, 2.16 m, the rest of the position is more intact, the fissures are moderately developed.

The geostress measurement adopts the original geostress measurement method of hydraulic fracturing, and the borehole is drilled vertically upward at the guide point of S17 in No.3 lane of No.18 coal in Baliancheng coal mine, the depth of the borehole opening is 632 m. The 3 measurement holes are 10 m apart, and 1 section of fracturing measurement is carried out in each hole. In the measured hole test, the rupture pressure, retensioning pressure and closure pressure in all measured holes can be clearly seen in each cycle, and the repeatability is relatively good. Therefore, the maximum horizontal principal stress and the minimum horizontal principal stress at each section could be calculated from the retension pressure and the closure pressure, respectively. Then the vertical stress (S_V) value of each section was obtained according to the thickness of the overlying rock layer and the rock bulk weight (2.5 g/cm³). The detailed results of the three holes are shown in ( Figure 3 ) and ( Table 2 ).

According to the data in Table 2 , the maximum horizontal principal stress (S_H) is 18.6 - 22.8 MPa, the minimum horizontal principal stress (S_h) is 10.22 - 11.68 MPa, and the vertical principal stress (S_V) is 15.47 MPa. It can be seen that the horizontal stress is dominant at this location, and S_H > S_V > S_h. The measured lateral pressure coefficient ranges from 1.20 to 1.47, with an error margin of 0 - 0.27, demonstrating negligible influence on the stability classification.

Annotation：¹P_b—In-situ rock breakdown pressure; ²P_r—Rupture surface reopening pressure; ³S_s—Rupture surface closing pressure;⁴S_h—Horizontal minimum principal stress; ⁵S_H—Horizontal maximum principal stress; ⁶S_v—Vertical principal stress; ⁷The vertical stress S_v is calculated by taking the bulk weight of the overlying rock as: 2.5 g/cm³.

Using borehole imaging technology, the significant rupture directions of the three test holes can be observed as shown in ( Figure 4 ), and the directions of the maximum horizontal principal stresses S_H are determined to be N45.85˚E, N29.69˚E and N39.32˚E, respectively.

Based on the engineering conditions of Baliancheng coal mine, FLAC3D was used to simulate the mining work of the 18 coal seam to obtain the evolution law of the mining stress field in the process of 18 coal seam back mining.

1) Modeling. The FLAC3D computer model was established with the top and bottom plate of the 18 coal seam in Baliancheng coal mine, and the model size was 400 m × 500 m × 282 m. The x-direction was used as the tendency of the working face, and the y-direction was the direction of the working face. The model contains 1,160,502 nodes and 1,128,370 hexahedral elements. The bottom boundary was fixed with displacement constraints, while the lateral boundaries (x and y directions) were assigned normal displacement constraints. A vertical stress of 9.725 MPa was applied to simulate the overburden pressure. The Mohr-Coulomb constitutive model was adopted with a unit weight of 2.5 kg/m³. Based on in-situ stress measurements, the lateral pressure coefficients were determined as 1.4 and 1.2 for the X and Y directions, respectively. The mechanical parameters of coal and rock strata were assigned according to laboratory test results, employing the Mohr-Coulomb failure criterion. The initial stress field was balanced to calibrate both the magnitude and orientation of stresses. Subsequent excavation simulations were performed to analyze the abutment pressure distribution around the working face. The simulated results were validated against field measurements to ensure accuracy.

During mining, the roof plate at the back of the working face collapses with the advancement of the working face, and goaf is filled with a large amount of bubbling rock, which is compacted under the influence of overburden loading, forming a support for the fissure zone. For this reason, when the numerical simulation is carried out, the bubbling zone after the hollowing is filled first, and the collapsed zone is filled with a double yield model.

2) Mining stress distribution law. Coal and rock in the natural state there is a certain stress state, basically in equilibrium, by mining disturbance, coal and rock stress redistribution, the formation of mining stress.18 coal seam working face maximum width of 190 m, excavation simulation after over the front mining stress cloud and stress curve shown in ( Figure 5 , Figure 6 ).

From the analysis of ( Figure 5 , Figure 6 ), it can be seen that after mining the working face of 18 seam, the vertical stress and horizontal stress in front of the working face show the trend of “increase-decrease”; the change rule of vertical stress is as follows: the vertical stress in the range of 0 - 20 m in front of the working face is gradually increasing, and the peak value of the vertical stress is 32.50 MPa, which is 2.71 times of the stress in the original rock; the vertical stress in the range of 20 - 130 m in front of the working face is gradually decreasing, the stress value is 12.81 - 32.50 MPa, 1.07 - 2.71 times of the original rock stress, of which the vertical stress in the range of 20 - 30 m in the overtopping is 18.46 - 32.50 MPa, 1.54 - 2.71 times of the original rock stress; The change rule of horizontal stress is as follows: overtopping 0 - 20 m range of horizontal stress gradually increases, and the peak stress value is 30.10 MPa, which is 1.79 times of the original rock stress; the horizontal stress in the range of 20 - 130 m ahead gradually decreases, and the stress value is 17.73 - 30.10 MPa, which is 1.05 - 1.79 times of the original rock stress. To summarize, the vertical stress in the working face of 18 seam increases more obviously in the range of overtopping, and the horizontal stress increases relatively small, and the mining stress affects more obviously in the range of 0 - 24 m in the overtopping working face.

1) Determine the classification index and evaluation method ( Wang, 2015 ). Drawing on the existing classification methods, synchronously consider the key factors such as top plate strength, coal seam strength, bottom plate strength, roadway depth, coal (rock) pillar width, mining influence coefficient, perimeter rock stability index, stress ratio and so on, and put forward the comprehensive index classification method, as shown in Equation (1), which is shown in ( Table 3 , Table 4 ).

$W_t=\frac{{\displaystyle \underset{i=1}{\overset{n}{\sum }}{W}_i}}{{\displaystyle \underset{i=1}{\overset{n}{\sum }}{W}_{i\max }}}$ (1)

The comprehensive index method formula was established in accordance with the rockburst risk classification standards specified in the “Interim Measures for the Identification of Rockburst-prone Mines” (Mine Safety [2023] No. 58) issued by the National Mine Safety Administration. The weighting coefficients for roadway depth and surrounding rock stability index were determined based on Reference ( Wang, 2015 ), while the weighting factors for roof strength, coal seam strength, floor strength, coal (rock) pillar width, mining influence coefficient, and stress ratio were derived from Reference ( Bai et al., 2020 ).

2) Classification results. According to the strength of the top and bottom plates and coal seams of the Baliancheng coal mine roadway, the depth of the roadway, the direct top initial fall step, the width of the coal guarding roadway, the mining influence coefficient, the average step of the direct top initial collapse, the stress ratio and other data, to determine the classification results of the stability of the surrounding rock of the roadway of the Baliancheng coal mine, and the results of the stability class of the surrounding rock of the 18-coal roadway of the Baliancheng coal mine are all medium-stable perimeter rock shown in ( Table 5 ).

In order to ensure safety and normal production, the roadway should maintain a certain stability, i.e., according to the requirements of safety regulations and technical regulations, the roadway should maintain the required shape and section size during its use. The classification method for the stability of the roadway surrounding rock in gently sloping and inclined coal seam back mining takes the strength of the roadway top and bottom plates and coal seams (σ_D, σ_d, σ_M), the depth of the roadway (H), the direct top initial fall distance (L), the ratio of the direct top thickness to the mining height (N), and the width of the coal protection roadway (X), which are the seven main factors influencing the stability of the perimeter rock of the roadway in the mining area, as classification indexes for the evaluation of the stability of the roadway. At present, China has formulated a gently sloping and inclined coal seam back to mining roadway rock stability classification scheme table, against the determination of the Baliancheng coal mine back to mining roadway rock stability classification results ( Table 6 ), the classification results are consistent with the comprehensive index method of roadway rock stability classification.

The perimeter rock stability class of the 18-coal roadway in Baliancheng coal mine belongs to Class III, i.e. medium stable perimeter rock. To enhance the stability of the surrounding rock, the original support scheme was optimized based on borehole camera observations of the roof. The results indicated that fracture concentration zones were predominantly within 2.16 m of the roadway roof. Consequently, the roof bolt length was increased from 2000 mm to 2200 mm. Field investigations revealed rock fall occurrences between bolts, prompting an increase in bolt density from 6 to 7 per row. Additionally, field measurements showed roof subsidence ranging between 170 - 190 mm. Considering the maximum elongation of anchor cables (3%) and the design principle that anchor cable length should be 1.5 times the roadway span, the anchor cable length was extended from 6500 mm to 7300 mm, and the number per row was increased from 2 to 3. For the rib support, the bolt length was extended from 1800 mm to 2000 mm, while the number and spacing remained unchanged. The optimized support scheme is illustrated in ( Figure 7 ).

Surface displacement observation of the roadway perimeter rock was carried out for 30 d under the present support scheme and the original support scheme, as shown in ( Figure 8 ).

From ( Figure 8 ), it can be seen that under the present support scheme, the maximum displacement of the top and bottom plates of the roadway is 164 mm, and the maximum displacement of the two gangs is 283 mm, and the overall deformation of the roadway is relatively reduced compared with that of the original support scheme, and the support effect of the roadway surrounding rock is good.