Study on the Pressure-Bearing Law of Backﬁlling Material Based on Three-Stage Strip Backﬁlling Mining

: During strip backﬁlling mining in coal mines, the backﬁlling material is the main support structure. Therefore, studying the pressure law of the backﬁlling material is essential for the safe and e ﬃ cient mining of coal resources. Based on research into strip backﬁlling mining at working face number 3216 of the Shanghe Coal Mine, and to smooth transition of overlying strata loads to the backﬁlling material, this study proposes a three-stage strip backﬁlling mining method. Based on thin-plate theory, an elastic thin-plate model, a reasonable spacing of strip mining is constructed, and the reasonable mining parameters of “mining 7 m to retain 8 m” at working face number 3216 of the Shanghe Coal Mine are determined. The law of backﬁlling pressure in three-stage strip backﬁlling mining is studied through numerical simulation and physical simulation experiments. The results show that ﬁeld measurement results are basically consistent with the experimental results and numerical simulation results. When three-stage strip backﬁlling mining is adopted, the stage-one backﬁlling material is the main bearing body to which the overlying rock load transfers smoothly and gradually, and the structure of the “overburden-coal pillar (or backﬁlling strip)” in the stope remains stable. In three-stage strip backﬁlling mining, the overlying rock load is ultimately transferred to the stage-one backﬁlling material, the stage-two backﬁlling material is the auxiliary bearing body, and the stage-three backﬁlling material mainly provides long-term stable lateral support for the stage-one backﬁlling material.


Introduction
The ecological environment in the northern Shaanxi province of China is fragile, and it belongs to the area of "coal mining with water conservation". Room and pillar mining (RPM), which is a simple mining method that involves less mining investment, was once a widely used mining technique for small and medium-sized coal mines in the northern Shaanxi province of China [1]. However, long-term creep of the remaining pillars in RPM is liable to cause large-scale instability of the bearing structure of overburden-pillar groups, resulting in a series of problems, such as tremors and surface subsidence [2][3][4][5]. At the same time, the recovery rate of the room pillar mining method is low. Strip backfilling mining (SBM) is superior to RPM in terms of recovery rate, surface subsidence control,

Selection of Mining and Backfilling Methods
The purpose of SBM is to replace solid coal with the BM. In the mining process, there should be a transition stage between the overlying rock load and BM. In other words, the overlying rock load should be transferred slowly and evenly to the BM. The research divides SBM into three stages to load the BM slowly and evenly.
The first stage is strip mining is shown in Figure 1. Here, adjacent strip faces are mined in turn, and strip pillars of certain sizes are set between the two strip goafs. The second stage is the replacement and recovery of even-numbered strip coal pillars and backward arrangement of the mining technology. The third stage is the replacement and recovery of odd-numbered strip coal pillars, with backward arrangement of the mining technology. After each strip mining stage, backfilling is carried out in accordance with the forward mode. should be transferred slowly and evenly to the BM. The research divides SBM into three stages to load the BM slowly and evenly. The first stage is strip mining is shown in Figure 1. Here, adjacent strip faces are mined in turn, and strip pillars of certain sizes are set between the two strip goafs. The second stage is the replacement and recovery of even-numbered strip coal pillars and backward arrangement of the mining technology. The third stage is the replacement and recovery of odd-numbered strip coal pillars, with backward arrangement of the mining technology. After each strip mining stage, backfilling is carried out in accordance with the forward mode.

Calculation of Strip Mining Parameters
While advancing along the strike of the strip mining face, the exposed direct roof stratum is a rectangular plate structure in space and has a fixed boundary condition on four sides. In elastic mechanics, the object enclosed by two parallel planes and cylinders perpendicular to these planes is called a plate. The distance t between the two planes is called the thickness of the plate, and the plane dividing t is called the middle plane of the plate. Considering that the thickness of the direct roof stratum is relatively small during strip mining in the Shanghe Coal Mine, and the area of the exposed roof stratum in the mining process is relatively large, the basic conditions for the study of thin plate theory are met [43].
Let the edge length of the thin rectangular plate formed by the exposed direct roof strata be 2a × 2b. Parameter 2a represents the length of the working face along an inclined direction (and has a fixed value), while parameter 2b represents the distance of the working face advancing along strike (and is variable). The area around the thin rectangular plate is in a fixed state. At the same time, to simplify the calculation and consider the near-horizontal coal seam, the influence of coal seam dip is not taken into account. A Cartesian coordinate system is established at the center of the middle plane of the plate, and q is assumed to be the vertical load on the main roof. The mechanical model of the thin rock plate is shown in Figure 2.
Assuming that the rock slab material is homogeneous, continuous, and identical, the loadbearing action of the thin plate can be analyzed by using the theory of small deflection of the elastic plate. In other words, the points on the normal line of the middle surface of the thin plate (before deformation) continue to exist on the same normal line of the elastic surface after deformation, and the distance between the points on the normal line remains unchanged, while the other stresses are separated. In comparison, the stress component z σ perpendicular to the mid-plane can be ignored. This model is aimed at shallow coal seam mining in northern Shaanxi province, China. When strip backfilling mining is adopted in this area, one strip after another will be formed. The roof above the strip is not bedded intensively, so this thin plate model can be used.

Calculation of Strip Mining Parameters
While advancing along the strike of the strip mining face, the exposed direct roof stratum is a rectangular plate structure in space and has a fixed boundary condition on four sides. In elastic mechanics, the object enclosed by two parallel planes and cylinders perpendicular to these planes is called a plate. The distance t between the two planes is called the thickness of the plate, and the plane dividing t is called the middle plane of the plate. Considering that the thickness of the direct roof stratum is relatively small during strip mining in the Shanghe Coal Mine, and the area of the exposed roof stratum in the mining process is relatively large, the basic conditions for the study of thin plate theory are met [43].
Let the edge length of the thin rectangular plate formed by the exposed direct roof strata be 2a × 2b. Parameter 2a represents the length of the working face along an inclined direction (and has a fixed value), while parameter 2b represents the distance of the working face advancing along strike (and is variable). The area around the thin rectangular plate is in a fixed state. At the same time, to simplify the calculation and consider the near-horizontal coal seam, the influence of coal seam dip is not taken into account. A Cartesian coordinate system is established at the center of the middle plane of the plate, and q is assumed to be the vertical load on the main roof. The mechanical model of the thin rock plate is shown in Figure 2.
Assuming that the rock slab material is homogeneous, continuous, and identical, the load-bearing action of the thin plate can be analyzed by using the theory of small deflection of the elastic plate. In other words, the points on the normal line of the middle surface of the thin plate (before deformation) continue to exist on the same normal line of the elastic surface after deformation, and the distance between the points on the normal line remains unchanged, while the other stresses are separated. In comparison, the stress component σ z perpendicular to the mid-plane can be ignored. This model is aimed at shallow coal seam mining in northern Shaanxi province, China. When strip backfilling mining is adopted in this area, one strip after another will be formed. The roof above the strip is not bedded intensively, so this thin plate model can be used. The strip-mining face starts to advance from the cutting hole, and the exposed roof stratum exists as a thin plate with a fixed support on all the sides. Its boundary condition is given as follows [44]: Under the action of load q, the thin plate deflects, and the deflection function of the plate is set as The deformation potential energy of the plates is given as follows: When Equation (4) is solved, the following is obtained: The total potential energy of the plate is According to the principle of minimum potential energy, the true displacement function ω of the thin plate should minimize Π . In short, the first derivative of Π is zero.
By substituting Equations (5)- (7) in Equation (8), the following results are obtained: The deflection expression is obtained by substituting C1 in Equation (3) as The strip-mining face starts to advance from the cutting hole, and the exposed roof stratum exists as a thin plate with a fixed support on all the sides. Its boundary condition is given as follows [44]: Under the action of load q, the thin plate deflects, and the deflection function of the plate is set as The deformation potential energy of the plates is given as follows: When Equation (4) is solved, the following is obtained: The potential energy of a normal load q N is The total potential energy of the plate is According to the principle of minimum potential energy, the true displacement function ω of the thin plate should minimize Π. In short, the first derivative of Π is zero.

of 16
The deflection expression is obtained by substituting C 1 in Equation (3) as where D is flexural rigidity of the plate and defined as D = Et 3 12(1−µ 2 ) . The stress component along the Y-axis of the thin plate can be expressed as Substituting Equations (10) and (9) in Equation (11) yields: where t is the plate thickness and µ is Poisson's ratio.
During the advancing process of the working face, the maximum tensile stress occurs at position x = 0, y = ±b, z = −t/2, meaning the middle of the long side (the length direction of the working face) of the thin stratum plate surface (the side of the back goaf). Along the y-axis direction, the value is With the advancing working face, tension cracks originate at the middle point of the long side of the upper surface of the thin plate's supporting end (along the length direction of the working face). Accordingly, σ max ≥ σ s (ultimate tensile strength of the rock plate) can be used as the criterion for determining the initial breakage of the rock plate, which can be expressed as follows: The load (q n ) 1 of overlying n strata on the direct roof can be calculated based on the following equation [45]: where E 1 , E 2 , . . . , E n is the elastic modulus of roof strata; h 1 , h 2 , . . . , h n is the thickness of roof strata; and γ 1 , γ 2 , · · · , γ n is the volume weight of roof strata. If (q n+1 ) 1 < (q n ) 1 , the load q imposed by the overlying nth stratum on the direct roof stratum can be calculated as (q n ) 1 .
The thickness of the immediate roof strata of Shanghe Coal Mine is 4.53 m, and the bulk density is 26.5 kN/m 3 , while the thickness of the feldspar sandstone is 24.51 m and the bulk density is 27.86 kN/m 3 . The load q 1 of the immediate roof stratum is given as The load action of the second layer on the first layer (q 2 ) 1 is as follows: As a result, the middle and coarse-grained feldspar sandstone in the second layer roof is separated from the roof below. The gray mudstone in the first layer only bears its own weight. At this point, the load value on the rock layer can be considered as (q 1 ) 1 = 120 kPa, which indicates that q = q 1 = 120 kPa.
Equation (14) is suitable for the shallowly buried coal seam, as the roof is not bedded intensively. According to the geological conditions of the Shanghe Coal Mine, this formula is applicable. Four parameters of the Shanghe Coal Mine are substituted in Equation (14). Among them, the length of working face 2a = 80 m, load q = 120 kPa on the direct roof, thickness of slab t = 2.2 m, and ultimate tensile strength of roof coal mudstone σ s = 900 kPa were used to calculate parameter b. Further, if b = 3.73 m, the strip width L = 2b = 7.46 m, and L = 7 m.
For water-conservation mining, mining should not destroy the aquifer structure. Therefore, according to Wilson's empirical formula for the width of strip coal pillars, we calculated the width B. The empirical formula is given as follows: where B is the coal pillar width given in m; L is the strip width of 7 m; M is the coal pillar height and H is 5.5 m; the coal seam depth is 115 m; F is the safety factor of strip coal pillar, with a value of 1.3. After substituting the data, B = 8.06 m, that is, the coal pillar width is 8 m.
The study of SBM has been carried out at working face number 3216 of the Shanghe Coal Mine. In view of its good geological conditions and stable roof, the strip mining parameters of "mining 7 m to retain 8 m" were adopted for safety reasons. [46]. When the density and geometric similarity constants are chosen, the stress similarity constant of the model is obtained. The major similarity coefficients were chosen as follows: Geometric similarity ratio α L = 100, bulk density ratio α γ = 1.53, and strength ratio α R = 153. Physical experimental materials were made of river sand as aggregate, calcium carbonate and gypsum as cementitious materials, and water in an appropriate proportion. The physical model was stratified according to the different strata lithology, and a layer of mica was placed among the layers to simulate the structural planes among layers. The physical experimental material ratio of working face number 3216 is listed in Table 1.

Construction of a Physical Model
The backfilling mining of working face number 3216 in the Shanghe Coal Mine was divided into three stages. In the first stage, the strip mining method of "mining 7 m to retain 8 m" was adopted. In the second stage, even-numbered strip pillars were recovered, and odd-numbered strip pillars were recovered during the third stage. According to the backfilling mining technology, the forward backfilling mode was adopted in all three stages. As shown in Figure 3.  The key of the experiment is to prepare the BM in the strip backfilling mining, according to the mining parameters of mining 7 m to retain 8 m. In this paper, aeolian sand (a kind of yellow sand peculiar to northern Shaanxi, China), fly ash, and cement were used as physical simulation BM. Through the strength, plasticity, and other aspects of testing, it was concluded that the ratio of BM suitable for the physical simulation experiment of Shanghe Coal Mine was aeolian sand: fly ash: cement = 15:2:1. The BM at this ratio can meet the bearing capacity required by the BM in the physical experiment. The prefabricated BM is shown in Figure 4. The BM was put into the strip goaf during SBM.  The key of the experiment is to prepare the BM in the strip backfilling mining, according to the mining parameters of mining 7 m to retain 8 m. In this paper, aeolian sand (a kind of yellow sand peculiar to northern Shaanxi, China), fly ash, and cement were used as physical simulation BM. Through the strength, plasticity, and other aspects of testing, it was concluded that the ratio of BM suitable for the physical simulation experiment of Shanghe Coal Mine was aeolian sand: fly ash: cement = 15:2:1. The BM at this ratio can meet the bearing capacity required by the BM in the physical experiment. The prefabricated BM is shown in Figure 4. The BM was put into the strip goaf during SBM. The key of the experiment is to prepare the BM in the strip backfilling mining, according to the mining parameters of mining 7 m to retain 8 m. In this paper, aeolian sand (a kind of yellow sand peculiar to northern Shaanxi, China), fly ash, and cement were used as physical simulation BM. Through the strength, plasticity, and other aspects of testing, it was concluded that the ratio of BM suitable for the physical simulation experiment of Shanghe Coal Mine was aeolian sand: fly ash: cement = 15:2:1. The BM at this ratio can meet the bearing capacity required by the BM in the physical experiment. The prefabricated BM is shown in Figure 4. The BM was put into the strip goaf during SBM.

Numerical Modeling
The numerical simulation model was established by combining geological engineering conditions with laboratory-measured physical and mechanical parameters of surrounding rocks. The boundary around the model is defined as a single constraint boundary, whereas the bottom boundary is defined as a full constraint boundary. The meshing model is shown in Figure 5, which consists of 231,000 units. To facilitate the study, the bedrock above is equivalent to 1.4 MPa uniform load. According to the geological engineering conditions and in situ stress data of the Shanghe Coal Mine, 4 MPa horizontal stress was applied around the model.
The Mohr-Coulomb constitutive relation was adopted in the numerical simulation model. Table 2 lists the physical and mechanical properties of rocks and BM used in the model. Among them, layer thickness was obtained based on the borehole histogram of Shanghe Coal Mine. The following parameters were measured in the laboratory: bulk density, Young's modulus (E), Poisson's ratio (µ), tensile strength, cohesion, and angle of internal. This was calculated by the following formula for bulk modulus (K) and shear modulus (G): In the numerical simulation model, the model is provided with a 20 m protective coal pillar and return air roadway. According to mining parameters of mining 7 m to retain 8 m and the three-stage SBM method, simulated mining was carried out.

Numerical Modeling
The numerical simulation model was established by combining geological engineering conditions with laboratory-measured physical and mechanical parameters of surrounding rocks. The boundary around the model is defined as a single constraint boundary, whereas the bottom boundary is defined as a full constraint boundary. The meshing model is shown in Figure 5, which consists of 231,000 units. To facilitate the study, the bedrock above is equivalent to 1.4 MPa uniform load. According to the geological engineering conditions and in situ stress data of the Shanghe Coal Mine, 4 MPa horizontal stress was applied around the model.
The Mohr-Coulomb constitutive relation was adopted in the numerical simulation model. Table  2 lists the physical and mechanical properties of rocks and BM used in the model. Among them, layer thickness was obtained based on the borehole histogram of Shanghe Coal Mine. The following parameters were measured in the laboratory: bulk density, Young's modulus (E), Poisson's ratio (μ), tensile strength, cohesion, and angle of internal. This was calculated by the following formula for bulk modulus (K) and shear modulus (G): In the numerical simulation model, the model is provided with a 20 m protective coal pillar and return air roadway. According to mining parameters of mining 7 m to retain 8 m and the three-stage SBM method, simulated mining was carried out.

Field Measurement Methods
To monitor the change in pressure in BM during SBM, a diaphragm pressure gauge was embedded into BM during strip goaf backfilling. The pressure gauge model was Bzy-50 made by Kejia special pressure gauge Co., Ltd., Wuxi, China (shown in Figure 6). To eliminate the influence of the boundary effect, the diaphragm of the pressure gauge was placed 2 m away from the return

Field Measurement Methods
To monitor the change in pressure in BM during SBM, a diaphragm pressure gauge was embedded into BM during strip goaf backfilling. The pressure gauge model was Bzy-50 made by Kejia special pressure gauge Co., Ltd., Wuxi, China (shown in Figure 6). To eliminate the influence of the boundary effect, the diaphragm of the pressure gauge was placed 2 m away from the return air roadway and 0.5 m away from the floor of the coal seam (see Figure 7). The buried pressure gauge from the field is shown in Figure 8. air roadway and 0.5 m away from the floor of the coal seam (see Figure 7). The buried pressure gauge from the field is shown in Figure 8.    Figure 9 shows that after the first stage of backfilling, pressure in the stage-one BM was generally less than 0.5 MPa. At this point, the stage-one BM and overlying strata gradually began to come into contact. After the second stage of backfilling, the pressure of the stage-one BM increased slightly (generally between 1.0 MPa and 1.5 MPa), and the overlying strata load transferred to stage-one BM. After the third air roadway and 0.5 m away from the floor of the coal seam (see Figure 7). The buried pressure gauge from the field is shown in Figure 8.    Figure 9 shows that after the first stage of backfilling, pressure in the stage-one BM was generally less than 0.5 MPa. At this point, the stage-one BM and overlying strata gradually began to come into contact. After the second stage of backfilling, the pressure of the stage-one BM increased slightly (generally between 1.0 MPa and 1.5 MPa), and the overlying strata load transferred to stage-one BM. After the third air roadway and 0.5 m away from the floor of the coal seam (see Figure 7). The buried pressure gauge from the field is shown in Figure 8.    Figure 9 shows that after the first stage of backfilling, pressure in the stage-one BM was generally less than 0.5 MPa. At this point, the stage-one BM and overlying strata gradually began to come into contact. After the second stage of backfilling, the pressure of the stage-one BM increased slightly (generally between 1.0 MPa and 1.5 MPa), and the overlying strata load transferred to stage-one BM. After the third  Figure 9 shows that after the first stage of backfilling, pressure in the stage-one BM was generally less than 0.5 MPa. At this point, the stage-one BM and overlying strata gradually began to come into contact. After the second stage of backfilling, the pressure of the stage-one BM increased slightly (generally between 1.0 MPa and 1.5 MPa), and the overlying strata load transferred to stage-one BM. After the third stage of backfilling, all the strip coal pillars were recovered. The pressure of the stage-one BM increased sharply to between 2.0 MPa and 4.0 MPa, and the average pressure in the BM was 2.76 MPa. At this stage, the bearing capacity of the stage-one BM reached its maximum. According to the entire mining and backfilling process of three-stage SBM, pressure in the BM generally presents in three areas, with the stress increasing and decreasing at both ends of the section, and the stress peaking in the middle of the section. The peak stress area occurs at the end of the third stage of backfilling when all the strip pillars have been recovered, and the overlying strata load is carried by the BM. stage of backfilling, all the strip coal pillars were recovered. The pressure of the stage-one BM increased sharply to between 2.0 MPa and 4.0 MPa, and the average pressure in the BM was 2.76 MPa. At this stage, the bearing capacity of the stage-one BM reached its maximum. According to the entire mining and backfilling process of three-stage SBM, pressure in the BM generally presents in three areas, with the stress increasing and decreasing at both ends of the section, and the stress peaking in the middle of the section. The peak stress area occurs at the end of the third stage of backfilling when all the strip pillars have been recovered, and the overlying strata load is carried by the BM.  Figure 10 shows that after the second stage of backfilling, pressure in the stage-two BM was generally less than 0.5 MPa, which was similar to that of the stage-one BM. After the third stage of backfilling, pressure in the stage-two BM slightly increased and was generally between 0.5 MPa and 1.0 MPa. The stage-two BM began to bear pressure slowly. When Figures 9 and 10 are compared, there is no apparent peak stress area in the entire backfilling and mining process of stage-two BM, indicating that stage-one BM bore most of the overlying strata pressure, while the stage-two BM further bore the pressure of the rock. With respect to the entire mining and backfilling process, the pressure increase in stage-two BM was more uniform and slower than that of the stage-one BM.

Numerical Modelling Results and Analysis
As shown in Figure 11, the average vertical stress of the first stage strip pillar was 3.5 MPa, while the maximum vertical stress was 4.3 MPa. The vertical stress of the coal pillar decreased by 0.7 MPa on average compared to that of unfilled strips. The average vertical stress of BM in the first stage was  Figure 10 shows that after the second stage of backfilling, pressure in the stage-two BM was generally less than 0.5 MPa, which was similar to that of the stage-one BM. After the third stage of backfilling, pressure in the stage-two BM slightly increased and was generally between 0.5 MPa and 1.0 MPa. The stage-two BM began to bear pressure slowly. When Figures 9 and 10 are compared, there is no apparent peak stress area in the entire backfilling and mining process of stage-two BM, indicating that stage-one BM bore most of the overlying strata pressure, while the stage-two BM further bore the pressure of the rock. With respect to the entire mining and backfilling process, the pressure increase in stage-two BM was more uniform and slower than that of the stage-one BM. stage of backfilling, all the strip coal pillars were recovered. The pressure of the stage-one BM increased sharply to between 2.0 MPa and 4.0 MPa, and the average pressure in the BM was 2.76 MPa. At this stage, the bearing capacity of the stage-one BM reached its maximum. According to the entire mining and backfilling process of three-stage SBM, pressure in the BM generally presents in three areas, with the stress increasing and decreasing at both ends of the section, and the stress peaking in the middle of the section. The peak stress area occurs at the end of the third stage of backfilling when all the strip pillars have been recovered, and the overlying strata load is carried by the BM.  Figure 10 shows that after the second stage of backfilling, pressure in the stage-two BM was generally less than 0.5 MPa, which was similar to that of the stage-one BM. After the third stage of backfilling, pressure in the stage-two BM slightly increased and was generally between 0.5 MPa and 1.0 MPa. The stage-two BM began to bear pressure slowly. When Figures 9 and 10 are compared, there is no apparent peak stress area in the entire backfilling and mining process of stage-two BM, indicating that stage-one BM bore most of the overlying strata pressure, while the stage-two BM further bore the pressure of the rock. With respect to the entire mining and backfilling process, the pressure increase in stage-two BM was more uniform and slower than that of the stage-one BM.

Numerical Modelling Results and Analysis
As shown in Figure 11, the average vertical stress of the first stage strip pillar was 3.5 MPa, while the maximum vertical stress was 4.3 MPa. The vertical stress of the coal pillar decreased by 0.7 MPa on average compared to that of unfilled strips. The average vertical stress of BM in the first stage was

Numerical Modelling Results and Analysis
As shown in Figure 11, the average vertical stress of the first stage strip pillar was 3.5 MPa, while the maximum vertical stress was 4.3 MPa. The vertical stress of the coal pillar decreased by 0.7 MPa on average compared to that of unfilled strips. The average vertical stress of BM in the first stage was 0.8 MPa, and the bearing strength of BM increased continuously. The BM in the initial stage was not loaded during the entire backfilling process. With the advance of backfilling work, the stress value of stage-one BM began to increase, which indicates that the roof load gradually transferred to the structure of "coal pillar-backfilling material", and although the roof load was released to a certain extent, the strip coal pillar remained the main bearing body of overlying rock load at this point.
Energies 2019, 12, x FOR PEER REVIEW 11 of 16 0.8 MPa, and the bearing strength of BM increased continuously. The BM in the initial stage was not loaded during the entire backfilling process. With the advance of backfilling work, the stress value of stage-one BM began to increase, which indicates that the roof load gradually transferred to the structure of "coal pillar-backfilling material", and although the roof load was released to a certain extent, the strip coal pillar remained the main bearing body of overlying rock load at this point. Figure 11. Simulation of stress evolution in the first stage of the backfilling process.
As shown in Figure 12, the second stage backfilling process showed that the stage-two BM had no bearing capacity, the average stress in the stage-one was 2.5 MPa, and the vertical stress increased significantly. The average stress of odd-numbered strip pillars was 4.0 MPa, and the stress growth of the pillars was not significant. This indicates that with continuous mining and backfilling, the roof load gradually transferred to the bearing system with stage-BM as the main support body. The roof load was further released to the BM, and the structure of "overburden-pillar-backfilling material" in the stope remained stable. As shown in Figure 13, the third stage of the backfilling process showed that the loads of the first, second, and third stages were 3.5 MPa, 1.0 MPa and 0.5 MPa, respectively. It can be seen that most of the roof load had been transferred to the stage-one BM, and the stress of the stage-two BM gradually increased, which plays an auxiliary role in supporting the roof. In the third stage, the stagethree BM was in an unloaded state. The analysis and simulation results show that when backfilling strip goaf, three continuous forms of high strength and low strength backfilling bodies are adopted, which can help the overburden to form a good structural bearing role and ensure normal development of the backfilling mining. As shown in Figure 12, the second stage backfilling process showed that the stage-two BM had no bearing capacity, the average stress in the stage-one was 2.5 MPa, and the vertical stress increased significantly. The average stress of odd-numbered strip pillars was 4.0 MPa, and the stress growth of the pillars was not significant. This indicates that with continuous mining and backfilling, the roof load gradually transferred to the bearing system with stage-BM as the main support body. The roof load was further released to the BM, and the structure of "overburden-pillar-backfilling material" in the stope remained stable.
Energies 2019, 12, x FOR PEER REVIEW 11 of 16 0.8 MPa, and the bearing strength of BM increased continuously. The BM in the initial stage was not loaded during the entire backfilling process. With the advance of backfilling work, the stress value of stage-one BM began to increase, which indicates that the roof load gradually transferred to the structure of "coal pillar-backfilling material", and although the roof load was released to a certain extent, the strip coal pillar remained the main bearing body of overlying rock load at this point. Figure 11. Simulation of stress evolution in the first stage of the backfilling process.
As shown in Figure 12, the second stage backfilling process showed that the stage-two BM had no bearing capacity, the average stress in the stage-one was 2.5 MPa, and the vertical stress increased significantly. The average stress of odd-numbered strip pillars was 4.0 MPa, and the stress growth of the pillars was not significant. This indicates that with continuous mining and backfilling, the roof load gradually transferred to the bearing system with stage-BM as the main support body. The roof load was further released to the BM, and the structure of "overburden-pillar-backfilling material" in the stope remained stable. As shown in Figure 13, the third stage of the backfilling process showed that the loads of the first, second, and third stages were 3.5 MPa, 1.0 MPa and 0.5 MPa, respectively. It can be seen that most of the roof load had been transferred to the stage-one BM, and the stress of the stage-two BM gradually increased, which plays an auxiliary role in supporting the roof. In the third stage, the stagethree BM was in an unloaded state. The analysis and simulation results show that when backfilling strip goaf, three continuous forms of high strength and low strength backfilling bodies are adopted, which can help the overburden to form a good structural bearing role and ensure normal development of the backfilling mining. As shown in Figure 13, the third stage of the backfilling process showed that the loads of the first, second, and third stages were 3.5 MPa, 1.0 MPa and 0.5 MPa, respectively. It can be seen that most of the roof load had been transferred to the stage-one BM, and the stress of the stage-two BM gradually increased, which plays an auxiliary role in supporting the roof. In the third stage, the stage-three BM was in an unloaded state. The analysis and simulation results show that when backfilling strip goaf, three continuous forms of high strength and low strength backfilling bodies are adopted, which can help the overburden to form a good structural bearing role and ensure normal development of the backfilling mining.

Analysis of Field Measurement Results in Field Backfill Materials
We can see from Figure 14 that, after the second and third stages of backfilling mining, the pressure in the stage-one BM increased significantly, with a pressure rise from 0.5 MPa to 2.50 MPa, and the pressure value rose by about five times. According to the pressure in the stage-two BM in Figure 15, after the third stage of backfilling, the average pressure in the stage-two BM increased from 0.5 MPa to 1.4 MPa, and the pressure increased by about three times. Figure 16 shows the pressure of the stage-three BM. The pressure in the stage-three BM was distributed between 0.3 MPa and 0.5 MPa. According to the pressure distribution of the whole BM after the completion of coal pillar mining, the overburden load transferred to the stage-one BM and also to the stage-two BM, causing the pressure of the stage-one and the stage-two BM to rise.

Analysis of Field Measurement Results in Field Backfill Materials
We can see from Figure 14 that, after the second and third stages of backfilling mining, the pressure in the stage-one BM increased significantly, with a pressure rise from 0.5 MPa to 2.50 MPa, and the pressure value rose by about five times. According to the pressure in the stage-two BM in Figure 15, after the third stage of backfilling, the average pressure in the stage-two BM increased from 0.5 MPa to 1.4 MPa, and the pressure increased by about three times. Figure 16 shows the pressure of the stage-three BM. The pressure in the stage-three BM was distributed between 0.3 MPa and 0.5 MPa. According to the pressure distribution of the whole BM after the completion of coal pillar mining, the overburden load transferred to the stage-one BM and also to the stage-two BM, causing the pressure of the stage-one and the stage-two BM to rise.

Analysis of Field Measurement Results in Field Backfill Materials
We can see from Figure 14 that, after the second and third stages of backfilling mining, the pressure in the stage-one BM increased significantly, with a pressure rise from 0.5 MPa to 2.50 MPa, and the pressure value rose by about five times. According to the pressure in the stage-two BM in Figure 15, after the third stage of backfilling, the average pressure in the stage-two BM increased from 0.5 MPa to 1.4 MPa, and the pressure increased by about three times. Figure 16 shows the pressure of the stage-three BM. The pressure in the stage-three BM was distributed between 0.3 MPa and 0.5 MPa. According to the pressure distribution of the whole BM after the completion of coal pillar mining, the overburden load transferred to the stage-one BM and also to the stage-two BM, causing the pressure of the stage-one and the stage-two BM to rise.

Analysis of Field Measurement Results in Field Backfill Materials
We can see from Figure 14 that, after the second and third stages of backfilling mining, the pressure in the stage-one BM increased significantly, with a pressure rise from 0.5 MPa to 2.50 MPa, and the pressure value rose by about five times. According to the pressure in the stage-two BM in Figure 15, after the third stage of backfilling, the average pressure in the stage-two BM increased from 0.5 MPa to 1.4 MPa, and the pressure increased by about three times. Figure 16 shows the pressure of the stage-three BM. The pressure in the stage-three BM was distributed between 0.3 MPa and 0.5 MPa. According to the pressure distribution of the whole BM after the completion of coal pillar mining, the overburden load transferred to the stage-one BM and also to the stage-two BM, causing the pressure of the stage-one and the stage-two BM to rise.

Comparison of Experimental Results with Numerical Simulation Results and Field Measurement Results
Physical simulation experimental results show that the stage-one BM pressure increased from 0.5 MPa to between 2.0 MPa and 4.0 MPa, and the peak stress area appeared in the middle of the mining area after the third stage of backfilling, which was the main bearing body of the entire backfilling process. Pressure in the stage-two BM was between 0.5 MPa and 1.0 MPa, which was the auxiliary BM, and there was no apparent peak stress area. The numerical simulation results show that the pressure of stage-one BM increased from 0.8 MPa to 3.5 MPa, and that of stage two and three BM were 1.0 MPa and 0.5 MPa, respectively. The overburden structure of the stope was kept stable by adopting three-stage SBM. Field measurement results are as follows: The average pressure of the stage-one BM was 2.45 MPa, while for stage-two BM, it was about 1.4 MPa, and for stage-three BM, it was between 0.3 MPa and 0.5 MPa. The field measurement results are basically consistent with the experimental results.
The two experimental results coincide and consistently reflect that in the three-stage SBM process, stage-one BM plays an important role for the entire backfilling process, bearing most of the overlying strata load; stage-two BM further disperses the overlying strata load and bears a smaller proportion of the pressure, and stage-three BM presents an unloaded state. This shows that the BM can form a good bearing structure in three-stage SBM.

Conclusions
The following conclusions can be made from this study: (1) Based on the thin-plate model of elasticity mechanics, the calculation model of the stripmining width was established. Reasonable strip mining parameters of working face number 3216 in the Shanghe Coal Mine were determined. The strip-mining width was calculated to be 7 m, and the pillar width was determined to be 8 m, based on Wilson's empirical formula.
(2) Physical simulation experimental results show that after the completion of three stages of SBM at working face 3216 of the Shanghe Coal Mine, the average pressure in the stage-one BM was 2.76 MPa, while in stage-two BM it was generally between 0.5 MPa and 1.0 MPa. The field measurement results were basically consistent with the experimental results.
(3) The numerical simulation results show that the overlying rock load can be transferred to the stage-BM smoothly and gradually by using the three-stage SBM method, and the structure of "overlying rock-coal pillar" (or backfilling strip) in the stope remains stable.
(4) In three-stage SBM, the overlying strata load is ultimately carried by stage-one BM, stagetwo BM is the auxiliary bearing body, and stage-three BM mainly provides a long-term lateral support force for stage-one BM. BM of different strengths can be selected in different backfilling mining stages.

Comparison of Experimental Results with Numerical Simulation Results and Field Measurement Results
Physical simulation experimental results show that the stage-one BM pressure increased from 0.5 MPa to between 2.0 MPa and 4.0 MPa, and the peak stress area appeared in the middle of the mining area after the third stage of backfilling, which was the main bearing body of the entire backfilling process. Pressure in the stage-two BM was between 0.5 MPa and 1.0 MPa, which was the auxiliary BM, and there was no apparent peak stress area. The numerical simulation results show that the pressure of stage-one BM increased from 0.8 MPa to 3.5 MPa, and that of stage two and three BM were 1.0 MPa and 0.5 MPa, respectively. The overburden structure of the stope was kept stable by adopting three-stage SBM. Field measurement results are as follows: The average pressure of the stage-one BM was 2.45 MPa, while for stage-two BM, it was about 1.4 MPa, and for stage-three BM, it was between 0.3 MPa and 0.5 MPa. The field measurement results are basically consistent with the experimental results.
The two experimental results coincide and consistently reflect that in the three-stage SBM process, stage-one BM plays an important role for the entire backfilling process, bearing most of the overlying strata load; stage-two BM further disperses the overlying strata load and bears a smaller proportion of the pressure, and stage-three BM presents an unloaded state. This shows that the BM can form a good bearing structure in three-stage SBM.