Stability Analysis of the Entry in a New Mining Approach Inﬂuenced by Roof Fracture Position

: Non-coal pillar mining with roadway formed automatically (RFANM) is a new mining approach, which demonstrates revolutionary signiﬁcance because it does not require making roadway before mining and coal pillar retaining. In order to explore the stability of the surrounding rock structure in RFANM, the deformation of the surrounding rock was theoretically analyzed and simulated based on three di ﬀ erent fracture positions of the main roof. It was concluded that reasonable control of temporary support strength in roadway is of great importance to control the deformation of the entry. The deformation process of surrounding rock under di ﬀ erent fracture positions in RFANM was simulated by using the Universal Discrete Element Code (UDEC). The results of the numerical simulation showed that the main roof was fractured at the solid coal side or gob side; the deformation of the roadway was small. The fracture condition of the main roof at the gob side required a higher e ﬀ ect of roof slitting or temporary support from the roadway. Through drilling and peeping at the retained roadway, it was judged that the main roof was broken inside the coal wall. Field monitoring results revealed that the deformation of the roadway can be e ﬀ ectively controlled.


Introduction
Longwall mining is the most common underground mining method in the world [1][2][3]. With the reduction of coal resources, there are problems in traditional underground coal mining, such as unaffordable driving costs, large surface disturbances, low coal-extraction rates, and the frequent occurrence of safety accidents. Such issues have also been addressed in the literature [4][5][6][7]. Following the development path of environment-friendly, safe, high-efficient, non-pillar mining technology, it has been widely used in coal mining. Gob-side entry retaining is a common practice of non-pillar mining in mine shafts, which significantly improves the coal recovery rate, reduces the roadway excavation, allows for the Y-type ventilation on the mining face, and addresses the gas accumulation at the corners of the working surface [8][9][10].
However, the gob-side entry retaining introduces various drawbacks as a result of coal (rock) pillar retaining, such as stress concentration in the surrounding rock and high risk of gas explosion in the gob zone. Non-coal pillar mining with roadway formed automatically (RFANM) is a new mining technology with no request required for excavating roadways in advance and with coal retaining pillar, The sections orthogonal to the mining direction shown in Figure 2 show that the technological process of the RFANM technique can be divided into the following steps: First, in working-face mining, the shearer cuts the coal while leaving the roadway space and forms the arc roadway wall on the side of the coal wall. Then, the roof of retained roadway is supported by constant resistance and large-deformation anchor cable [24]. In the predefined position above the slit drill, the roof is sliced by the bidirectional, energy-concentrated tension-blasting method, which cuts off the stress transfer between the gob and the roadway roof. The gob roof is automatically caved along the slit face under the action of mine pressure. At the same time, the fallen gangue fills the gob. A gravel gang is formed on the side of the gob under the action of roof-cutting support, U-shaped steel, and other gangue support constructions. Using the RFANM method, non-pillar mining without advancement is successfully implemented.   The sections orthogonal to the mining direction shown in Figure 2 show that the technological process of the RFANM technique can be divided into the following steps: First, in working-face mining, the shearer cuts the coal while leaving the roadway space and forms the arc roadway wall on the side of the coal wall. Then, the roof of retained roadway is supported by constant resistance and large-deformation anchor cable [24]. In the predefined position above the slit drill, the roof is sliced by the bidirectional, energy-concentrated tension-blasting method, which cuts off the stress transfer between the gob and the roadway roof. The gob roof is automatically caved along the slit face under the action of mine pressure. At the same time, the fallen gangue fills the gob. A gravel gang is formed on the side of the gob under the action of roof-cutting support, U-shaped steel, and other gangue support constructions. Using the RFANM method, non-pillar mining without advancement is successfully implemented.  The sections orthogonal to the mining direction shown in Figure 2 show that the technological process of the RFANM technique can be divided into the following steps: First, in working-face mining, the shearer cuts the coal while leaving the roadway space and forms the arc roadway wall on the side of the coal wall. Then, the roof of retained roadway is supported by constant resistance and large-deformation anchor cable [24]. In the predefined position above the slit drill, the roof is sliced by the bidirectional, energy-concentrated tension-blasting method, which cuts off the stress transfer between the gob and the roadway roof. The gob roof is automatically caved along the slit face under the action of mine pressure. At the same time, the fallen gangue fills the gob. A gravel gang is formed on the side of the gob under the action of roof-cutting support, U-shaped steel, and other gangue support constructions. Using the RFANM method, non-pillar mining without advancement is successfully implemented.

Deformation Characteristics of Roof Strata
With the advance of the working face and progressive placement of the hydraulic support accordingly, the roof strata lose their supporting capability, and the immediate roof collapses irregularly, leading to volume expansions, and therefore, filling up of the gob zone [25,26]. As the directional cutting is performed on the roadway roof where stress transfers between the roadway roof and the immediate gob roof, and a portion of the main roof was cut off, all rocks in the vicinity of the cutting range would cave down. Roof cutting forces the roadway roof to form a short beam structure, which is considered the "equivalent immediate roof" here.
As the working face continues to advance, when the hanging length of the main roof increases to the limit, the main roof breaks and sinks, resulting in periodic rock pressure appearance [27]. The gob-side main roof brakes into block A, B, and C, which engage with each other to form the masonry beam structure. Block B revolves and sinks; the end stretching to the gob side was the earliest to contact the gangue, tending to stabilize as the gangue is gradually compacted; therefore, forming a temporarily-balanced rock beam.
After the rock mass forms by the fracture of the main roof stabilizing, a number of rock groups in the overlying strata begin to separate, break, rotate, and cave down. In this stage, the roadway deformation is characterized by parallel subsidence. In the stable section of the roadway which lags behind a certain distance of the working face, the roof-cutting support in the roadway can be retracted and recycled.

Theoretical Model for Fracture of the Main Roof
During the roof displacement, the stage of the main roof fracture and the sag is the most intense stage of the roof displacement. Therefore, the theoretical model for fracture of the main roof in RFANM was established with reference to the roof motion in this stage. There are three possible fracture sites for the main roof in RFANM: solid coal side, above the roadway, and gob side.
According to the principle of RFANM and the characteristics of roof motion, the following assumptions of the theoretical model for fracture of the main roof in RFANM were made: (1) The roadway roof rotates and settles around the critical site of the solid, coal-side elastoplastic zone as a rotating axis; (2) the deformation range of key blocks is small, and the weights of strata which cannot form a stable masonry structure above the main roof act uniformly on the key blocks; (3) all roofs in the gob-side cutting range will cave; the supporting force against key blocks is uniformly loaded. (4) Because the angle between the blasting slit angle of the roof and the vertical direction is usually small, the influence of the angle of the slit on the force of the key block is negligible; (5) the curved lane is approximately as a straight lane; (6) the filling effect of gangues is ideal. After the main roof breaks, the rotating angle of block B is approximately zero, so that the horizontal thrust between blocks is negligible. (7) The horizontal shear stress between rock strata is negligible.

Coal-Side Fracture of the Main Roof
When the strength of temporary support for a roadway is insufficient or the support is provided late, the main roof breaks at the site with a length of x 0 from the solid coal elastoplastic zone of the roadway. One end of the key block B was extended beyond the solid coal side, and the other end was controlled by the gangue support in the gob and temporary support in the roadway. The free-body diagram analysis for the case of fracture of the main roof at the solid coal side was proposed, as shown in Figure 3. The equations for the width of the solid, coal-side stress limit equilibrium zone, x0, and the supporting force of the coal body against the roof, 0 σ , [28] are given by Equation (1): where A is the thickness of the coal seam in m; λ is the lateral pressure coefficient; c, is the interfacial cohesion between the boundary of coal seam and the roof and floor strata in MPa, and φ is the internal friction angle in deg; k is the stress concentration factor; γ is the average bulk density of the overlying strata in kN/m 3 ; H is the buried depth of the roadway in m; and P is solid coal support strength in MPa.
In the process of periodic roof collapse, the roof of a gob is in the state of three-points fixed and one-point free. The equation for the fracture length of key blocks is: where S is the length of the mining face and l is the periodic weighting length. According to Figure 3, blocks B and C were isolated using a static balance method for stress analysis. The simplified force equilibrium of key blocks is shown in Figure 4.  For block C, as 0 y F  = , we get As Simultaneously with Equations (4) and (5), we get The equations for the width of the solid, coal-side stress limit equilibrium zone, x 0 , and the supporting force of the coal body against the roof, σ 0 , [28] are given by Equation (1): where A is the thickness of the coal seam in m; λ is the lateral pressure coefficient; c, is the interfacial cohesion between the boundary of coal seam and the roof and floor strata in MPa, and ϕ is the internal friction angle in deg; k is the stress concentration factor; γ is the average bulk density of the overlying strata in kN/m 3 ; H is the buried depth of the roadway in m; and P is solid coal support strength in MPa.
In the process of periodic roof collapse, the roof of a gob is in the state of three-points fixed and one-point free. The equation for the fracture length of key blocks is: where S is the length of the mining face and l is the periodic weighting length. According to Figure 3, blocks B and C were isolated using a static balance method for stress analysis. The simplified force equilibrium of key blocks is shown in Figure 4. F y is the resultant force in the Y-axis direction of the key block and F y = 0 in the equilibrium state. M A is the total bending moment of key block B at point A' and M B is the total bending moment of key block C at point B'.
(a) Stress of key block C (b) Stress of key block B For block B, as 0 and where q0 is the average load of the immediate roof, in kN/m; and MA is the residual bending moment of key block, in kN·m. And as P1 = q3 × a, the temporary support strength in the roadway can be obtained by combining Equations (6)-(9).

Fracture Analysis of the Main Roof above the Roadway
When the main roof above the roadway is caved in, so block B is supported by the roadway support body and the gob gangue, the roof motion has the most obvious influence on the roadway deformation. When the crack development of the roadway roof is high or the temporary support in the roadway is provided late, the fracture site of the main roof appears above the roadway. Based on those assumptions, the theoretical calculation model for the fracture above the roadway of the main roof in RFANM was established, as shown in Figure 5. For block C, as F y = 0, we get As M B = 0, we get Simultaneously with Equations (4) and (5), we get For block B, as F y = 0, M A = 0, then and where q 0 is the average load of the immediate roof, in kN/m; and M A is the residual bending moment of key block, in kN·m. And as P 1 = q 3 × a, the temporary support strength in the roadway can be obtained by combining Equations (6)- (9).

Fracture Analysis of the Main Roof above the Roadway
When the main roof above the roadway is caved in, so block B is supported by the roadway support body and the gob gangue, the roof motion has the most obvious influence on the roadway deformation. When the crack development of the roadway roof is high or the temporary support in the roadway is provided late, the fracture site of the main roof appears above the roadway. Based on those assumptions, the theoretical calculation model for the fracture above the roadway of the main roof in RFANM was established, as shown in Figure 5. Sustainability 2019, 11, x FOR PEER REVIEW 7 of 15 Coal wall

Main roof
Floor Immediate roof The stress analysis of blocks B and C was performed using the static balance method. The stress on block C was the same as that under the gob-side fracture condition. Free body diagram of block B is presented in Figure 6. As where ' a is the distance from the fracture site of the roof to the edge of the gob-side roadway, in m. Simultaneously with Equations (6), (7), and (11), the temporary support strength in the roadway can be obtained.

Fracture Analysis of the Main Roof at the Gob Side
When the theoretical properties of the equivalent immediate roof manifest sufficient stand-up time or the temporary support strength in the roadway was large, the main roof broke along the slotting line. Under this fracture circumstance, key block B was stabilized with the gangue support. The length of the cantilever length above the roadway was less. The stability of the roadway formed was provided by the supporting force of solid coal and the support in the roadway. The theoretical model for the main roof fractures at the gob side was obtained, as shown in Figure 7. The simplified force equilibrium of key blocks A is shown in Figure 8. The stress analysis of blocks B and C was performed using the static balance method. The stress on block C was the same as that under the gob-side fracture condition. Free body diagram of block B is presented in Figure 6.
Coal wall

Main roof
Floor Immediate roof The stress analysis of blocks B and C was performed using the static balance method. The stress on block C was the same as that under the gob-side fracture condition. Free body diagram of block B is presented in Figure 6. As where ' a is the distance from the fracture site of the roof to the edge of the gob-side roadway, in m. Simultaneously with Equations (6), (7), and (11), the temporary support strength in the roadway can be obtained.

Fracture Analysis of the Main Roof at the Gob Side
When the theoretical properties of the equivalent immediate roof manifest sufficient stand-up time or the temporary support strength in the roadway was large, the main roof broke along the slotting line. Under this fracture circumstance, key block B was stabilized with the gangue support. The length of the cantilever length above the roadway was less. The stability of the roadway formed was provided by the supporting force of solid coal and the support in the roadway. The theoretical model for the main roof fractures at the gob side was obtained, as shown in Figure 7. The simplified force equilibrium of key blocks A is shown in Figure 8. As M A = 0, then where a is the distance from the fracture site of the roof to the edge of the gob-side roadway, in m. Simultaneously with Equations (6), (7), and (11), the temporary support strength in the roadway can be obtained.

Fracture Analysis of the Main Roof at the Gob Side
When the theoretical properties of the equivalent immediate roof manifest sufficient stand-up time or the temporary support strength in the roadway was large, the main roof broke along the slotting line. Under this fracture circumstance, key block B was stabilized with the gangue support. The length of the cantilever length above the roadway was less. The stability of the roadway formed was provided by the supporting force of solid coal and the support in the roadway. The theoretical model for the main roof fractures at the gob side was obtained, as shown in Figure 7. The simplified force equilibrium of key blocks A is shown in Figure 8. For block B, as 0 Simultaneously with the above equations, use: ( 2 )/2 0 Simultaneously with Equations (15) and (16), solve the temporary support strength in the roadway as: For convenience of simplification, we assume LA = LB; then 1 3  (10) and (17), obviously, the temporary support strength in the roadway when the fracture site of the main roof is at the solid coal side is greater than that required at the gob side; namely P1 > P3. The fracture occurrs above the roadway; one end of block B is on the roadway, and there is no supporting force of the solid coal. Furthermore, comparing Equations (10) and (12), we know that the temporary support For block B, as 0 Simultaneously with the above equations, use: Simultaneously with Equations (15) and (16), solve the temporary support strength in the roadway as: For convenience of simplification, we assume LA = LB; then 1 3  (10) and (17), obviously, the temporary support strength in the roadway when the fracture site of the main roof is at the solid coal side is greater than that required at the gob side; namely P1 > P3. The fracture occurrs above the roadway; one end of block B is on the roadway, and there is no supporting force of the solid coal. Furthermore, comparing Equations (10) and (12), we know that the temporary support For block B, as F y = 0, M A = 0, Simultaneously with the above equations, use: As M O = 0, then Simultaneously with Equations (15) and (16), solve the temporary support strength in the roadway as: For convenience of simplification, we assume L A = L B ; then P 1 − P 3 = 2(M 0 − M A )/(a + 2x 0 ) is obtained. Meanwhile, the residual bending moment of the beam fracture, M A , is much less than the bending moment occurring in the main roof, M O . By comparing the theoretical models for the fracture of the main roof at the gob side and at the solid coal side and Equations (10) and (17), obviously, the temporary support strength in the roadway when the fracture site of the main roof is at the solid coal side is greater than that required at the gob side; namely P 1 > P 3 . The fracture occurrs above the roadway; one end of block B is on the roadway, and there is no supporting force of the solid coal. Furthermore, comparing Equations (10) and (12), we know that the temporary support strength in the roadway when the fracture site of the main roof is at the coal side is greater than that required above the roadway; namely P 1 < P 2 . Therefore, it was concluded that when the temporary support residence in the roadway P > P 2 , the fracture site of the main roof outside the roadway is at the gob side or the solid coal side; when P < P 2 , the fracture of the main roof above the roadway leads to failure.

Numerical Modeling
As the lateral main roof may has three different fracture sites, the UDEC 2D model for the S1201-II mining face in Ningtiaota Coal Mine was established to simulate the movement of key rock blocks when the fracture position of main roof is on the side of coal wall, above the roadway and side of the gob.
The length and width of the model were 200 and 70 m, respectively. The load applied on the upper boundary was considered equal to the maximum buried depth of 170 m. The constitutive model of rock mass used was the Mohr-Coulomb model. The joint area contact-Coulomb slip model was used in the joint-constitutive model. The bottom boundary speed of the model was zero. And the horizontal direction of the left and right borders was limited to zero. The fracture sites in the main roof were x 0 = 4 m away from the curved lane (the width of the solid coal-side stress limit equilibrium zone), right above the roadway, and 2 m away from the roadway (right above the slot top). The numerical model was as shown in Figure 9. Cable elements were used to simulate constant resistance anchor cable support. In addition, supporting structural elements were used to simulate temporary support in the roadway. Physical properties of the rock strata obtained from the drilling cores are presented in Table 1 [29].
Sustainability 2019, 11, x FOR PEER REVIEW 9 of 15 strength in the roadway when the fracture site of the main roof is at the coal side is greater than that required above the roadway; namely P1 < P2. Therefore, it was concluded that when the temporary support residence in the roadway P > P2, the fracture site of the main roof outside the roadway is at the gob side or the solid coal side; when P < P2, the fracture of the main roof above the roadway leads to failure.

Numerical Modeling
As the lateral main roof may has three different fracture sites, the UDEC 2D model for the S1201-II mining face in Ningtiaota Coal Mine was established to simulate the movement of key rock blocks when the fracture position of main roof is on the side of coal wall, above the roadway and side of the gob.
The length and width of the model were 200 and 70 m, respectively. The load applied on the upper boundary was considered equal to the maximum buried depth of 170 m. The constitutive model of rock mass used was the Mohr-Coulomb model. The joint area contact-Coulomb slip model was used in the joint-constitutive model. The bottom boundary speed of the model was zero. And the horizontal direction of the left and right borders was limited to zero. The fracture sites in the main roof were x0 = 4 m away from the curved lane (the width of the solid coal-side stress limit equilibrium zone), right above the roadway, and 2 m away from the roadway (right above the slot top). The numerical model was as shown in Figure 9. Cable elements were used to simulate constant resistance anchor cable support. In addition, supporting structural elements were used to simulate temporary support in the roadway. Physical properties of the rock strata obtained from the drilling cores are presented in Table 1 [29].

Results and Discussions of Numerical Simulation
The caving forms of the main roof at different fracture sites are shown in Figure 10. When the main roof broke above the roadway, the extent of roadway deformation was greater than that in the other circumstances. When the main roof broke at the gob side, gangues mainly provided supporting forces against block B and the roadway was subjected to less pressure and the least deformation.

Results and Discussions of Numerical Simulation
The caving forms of the main roof at different fracture sites are shown in Figure 10. When the main roof broke above the roadway, the extent of roadway deformation was greater than that in the other circumstances. When the main roof broke at the gob side, gangues mainly provided supporting forces against block B and the roadway was subjected to less pressure and the least deformation. A surveying line equal to the width of the roadway was provided on the surface above the roadway to obtain the extent of lateral roof deformation of the roadway at different fracture sites, as shown in Figure 11. The extent of roadway roof settling near the gob is obviously higher than that at any other site. The maximum roof settling values at the solid coal side, above the roadway, and at the gob side were 130 mm, 230 mm, and 93 mm, respectively. The roadway deformation and failure when the fracture of the main roof occurred above the roadway were more significant than those of any other circumstance. A surveying line equal to the width of the roadway was provided on the surface above the roadway to obtain the extent of lateral roof deformation of the roadway at different fracture sites, as shown in Figure 11. The extent of roadway roof settling near the gob is obviously higher than that at any other site. The maximum roof settling values at the solid coal side, above the roadway, and at the gob side were 130 mm, 230 mm, and 93 mm, respectively. The roadway deformation and failure when the fracture of the main roof occurred above the roadway were more significant than those of any other circumstance. To gob side distance(m) Roof-to-floor convergence(mm) Coal side Gob side Above the roadway Figure 11. Roof migration of roadway with different fracture levels.
The variation curves of the settling observed in the roadway roof at the gob side, in the middle of the roadway, and at the curved coal side over time are shown in Figure 12a-c, respectively. The coal extraction from the mining face was completed; the deformation process of the main roof at the three fracture sites was divided into three stages-higher velocity in the early stage transitioning into a stabilized stage gradually. The deformation under the gob side fracture condition was concentrated in the early stage, and then slowly deformed until stabilized. The fracture process above the roadway was based on the deformation in former stages; the velocity of roof settling increased until the maximum values under the three fracture conditions were reached, respectively.

Formatting of Mathematical Components
The coal seam extracted on the test face was the middle section of Yanan Formation of Jurassic Series. The coal seam thickness was 3.81-4.35 m, the average thickness was 4.11 m, and the buried depth was 115-170 m. The occurrence of the coal seam was stable and its dig angle was approximately horizontal. The dip length and strike length of the working face were 280 and 2344 m, respectively. The position and the layout of the S1201-II working face are shown in Figure 13. The variation curves of the settling observed in the roadway roof at the gob side, in the middle of the roadway, and at the curved coal side over time are shown in Figure 12a-c, respectively. The coal extraction from the mining face was completed; the deformation process of the main roof at the three fracture sites was divided into three stages-higher velocity in the early stage transitioning into a stabilized stage gradually. The deformation under the gob side fracture condition was concentrated in the early stage, and then slowly deformed until stabilized. The fracture process above the roadway was based on the deformation in former stages; the velocity of roof settling increased until the maximum values under the three fracture conditions were reached, respectively. To gob side distance(m) Roof-to-floor convergence(mm) Coal side Gob side Above the roadway Figure 11. Roof migration of roadway with different fracture levels.
The variation curves of the settling observed in the roadway roof at the gob side, in the middle of the roadway, and at the curved coal side over time are shown in Figure 12a-c, respectively. The coal extraction from the mining face was completed; the deformation process of the main roof at the three fracture sites was divided into three stages-higher velocity in the early stage transitioning into a stabilized stage gradually. The deformation under the gob side fracture condition was concentrated in the early stage, and then slowly deformed until stabilized. The fracture process above the roadway was based on the deformation in former stages; the velocity of roof settling increased until the maximum values under the three fracture conditions were reached, respectively.

Formatting of Mathematical Components
The coal seam extracted on the test face was the middle section of Yanan Formation of Jurassic Series. The coal seam thickness was 3.81-4.35 m, the average thickness was 4.11 m, and the buried depth was 115-170 m. The occurrence of the coal seam was stable and its dig angle was approximately horizontal. The dip length and strike length of the working face were 280 and 2344 m, respectively. The position and the layout of the S1201-II working face are shown in Figure 13.

Formatting of Mathematical Components
The coal seam extracted on the test face was the middle section of Yanan Formation of Jurassic Series. The coal seam thickness was 3.81-4.35 m, the average thickness was 4.11 m, and the buried depth was 115-170 m. The occurrence of the coal seam was stable and its dig angle was approximately horizontal. The dip length and strike length of the working face were 280 and 2344 m, respectively. The position and the layout of the S1201-II working face are shown in Figure 13.

Roof Control and Roadway Formation Effect
Generally, the probability that the main roof breaks above the roadway is smaller than those for the solid coal side and the gob side, but its threat to the roadway stability is often the highest. During the overburdening strata movement, the temporary support in the roadway should not only provide a large supporting force to resist the deformation, but also functions as "yielding" to deal with the "given deformation" of the main roof. Cutting roof support is the main temporary support for the roadway. The support strength of roof-cutting support has an important influence on the fracture position on a basic roof and plays an important role in roadway stability.
RFANM follows the mining face closely to provide active support against the roadway roof with constant resistance and a large deformation anchor cable. The constant-resistance anchor cable is mounted before the surrounding rock of the roadway becomes deformed, so as to avoid the bed separation from the roadway roof. The support provided by the constant-resistance anchor cable enhances the anti-breaking capacity of the equivalent immediate roof, so that the fracture site of the lower roof tends to approach the gob side, which is beneficial to avoid the fracture of the main roof above the roadway.
By designing the slotting parameters, the gangues in the gob can achieve better effects of cracking-expansion and filling, which alleviate the impact of the intense motion of key blocks on the stability of the surrounding rock. Subject to the supporting force of the gangues in the gob and the temporary support in the roadway, the fracture line of the main roof shifts to the depth of the solid coal side or to the gob. This is significant for the stabilization of the surrounding rock of the roadway.
The variation curve of the roof and floor displacement with the advance of the mining face, as shown in Figure 14, was drawn by monitoring the roof displacement and selecting the typical measurement points with timely distribution and the longest monitoring duration at 930 m from the open-off cut. The variation of the roof and floor displacement was similar to that of the fracture of the main roof at the solid coal side and the gob side. The maximum roof and floor displacement at the gob side was 128.1 mm, which was approximately equal to the maximum deformation at the solid coal side. It was initially concluded that the main roof in the formation section broke at the solid coal side. The main roof of the mining face was medium-grained sandstone with a thickness of 5.4-21.5 m, a medium sorting behavior, and large cross-bedding. The immediate roof was siltstone with a thickness of 0.78-4.05 m. The immediate floor was siltstone with a thickness of 1.8-16.3 m, sandwiched and interbedded with a thin layer of fine-grained sandstone; the cleavage is shown. The main roof was fine-grained sandstone with a thickness of 3.2-19.6 m and wavy bedding.

Roof Control and Roadway Formation Effect
Generally, the probability that the main roof breaks above the roadway is smaller than those for the solid coal side and the gob side, but its threat to the roadway stability is often the highest. During the overburdening strata movement, the temporary support in the roadway should not only provide a large supporting force to resist the deformation, but also functions as "yielding" to deal with the "given deformation" of the main roof. Cutting roof support is the main temporary support for the roadway. The support strength of roof-cutting support has an important influence on the fracture position on a basic roof and plays an important role in roadway stability.
RFANM follows the mining face closely to provide active support against the roadway roof with constant resistance and a large deformation anchor cable. The constant-resistance anchor cable is mounted before the surrounding rock of the roadway becomes deformed, so as to avoid the bed separation from the roadway roof. The support provided by the constant-resistance anchor cable enhances the anti-breaking capacity of the equivalent immediate roof, so that the fracture site of the lower roof tends to approach the gob side, which is beneficial to avoid the fracture of the main roof above the roadway.
By designing the slotting parameters, the gangues in the gob can achieve better effects of cracking-expansion and filling, which alleviate the impact of the intense motion of key blocks on the stability of the surrounding rock. Subject to the supporting force of the gangues in the gob and the temporary support in the roadway, the fracture line of the main roof shifts to the depth of the solid coal side or to the gob. This is significant for the stabilization of the surrounding rock of the roadway.
The variation curve of the roof and floor displacement with the advance of the mining face, as shown in Figure 14, was drawn by monitoring the roof displacement and selecting the typical measurement points with timely distribution and the longest monitoring duration at 930 m from the open-off cut. The variation of the roof and floor displacement was similar to that of the fracture of the main roof at the solid coal side and the gob side. The maximum roof and floor displacement at the gob side was 128.1 mm, which was approximately equal to the maximum deformation at the solid coal side. It was initially concluded that the main roof in the formation section broke at the solid coal side. The roadway roof was drilled and peeped using a mine borescope. Figure 15 shows the peeping result at 930 m from the open-off cut and 10 cm from the coal wall with the drilling angles of 60° and 70°, respectively. It was observed that the peeping result of the fracture site of the main roof in this entry retaining section 3.6 m from the coal wall was basically consistent with the field displacement monitoring and simulation result. The field roadway formation effect is shown in Figure 16.  The roadway roof was drilled and peeped using a mine borescope. Figure 15 shows the peeping result at 930 m from the open-off cut and 10 cm from the coal wall with the drilling angles of 60 • and 70 • , respectively. It was observed that the peeping result of the fracture site of the main roof in this entry retaining section 3.6 m from the coal wall was basically consistent with the field displacement monitoring and simulation result. The field roadway formation effect is shown in Figure 16. The roadway roof was drilled and peeped using a mine borescope. Figure 15 shows the peeping result at 930 m from the open-off cut and 10 cm from the coal wall with the drilling angles of 60° and 70°, respectively. It was observed that the peeping result of the fracture site of the main roof in this entry retaining section 3.6 m from the coal wall was basically consistent with the field displacement monitoring and simulation result. The field roadway formation effect is shown in Figure 16.

Conclusions
In this study, the theoretical models for different fracture sites of the main roof were established, the equations for temporary support strength in the roadway when the main roof was at three fracture sites were derived, respectively, the critical value for the temporary support strength in the roadway was thus obtained, which offered theoretical evidence for the support design in RFANM.
We discovered that the fracture observed in the main roof at different sites was closely related to the stability of the surrounding rock of the roadway. The required temporary support in the roadway varied from site to site. When the main roof broke above the roadway, the temporary support strength required for the roadway to maintain stability was the largest. Thus, during the mining working in RFANM, the main roof was prevented from breaking above the roadway.
The deformation of key blocks at three fracture sites were simulated using the UDEC. The comparison of displacement variation of the roadway roof indicates that the roadway deformation is smaller when the main roof fractures at the solid coal side and the gob side.
The industrial test for RFANM was conducted on the S1201-IImining face, deformation of the roadway can be effectively controlled by using roof cutting support, constant resistance anchor cable support, resulting in good formation effect, better social and economic benefits.

Conflicts of Interest:
The authors declare no conflict of interest.