Research on Rock Movement Control Method in Underground Mining Based on the Action of Granular Lateral Pressure

Lateral pressure provided by caving rocks plays an important role in keeping the stability of the sidewall rock mass and controlling the rock movement. Taking Xilinhot Fluorite Mine as an engineering background, we studied the control mechanism of the granular media lateral pressure on the rock movement by combining physical experiments, theoretical analysis, and field monitoring. The results indicated that with the increase of the granular media depth, the lateral pressure provided by the static granular media presented two stages of rapid growth and slow growth. The lateral pressure on the footwall was negatively correlated with the dip angle of the orebody, and the lateral pressure on the hanging wall was positively correlated with the dip angle of the orebody. Ore drawing only caused the granular media lateral pressure near the drawing point to decrease. As the granular media moved, the granular media lateral pressure outside the pressure reduction area further increased, and this part of the granular media would play a key role in limiting surface rock movement. On this basis, the improved calculation formula of GLMP was given, which could be better applied to calculate the GLMP and critical granular media column with different dip angles. The action mechanism of the GLMP in the control of the surface rock movement was clarified. The granular body in the collapse pit was mainly composed of three parts: the bottom loose granular media column, the middle compaction granular media column, and the upper critical granular media column. The critical granular media column was the most important area to control surface subsidence and rock movement. Finally, through field practice, after effective filling of the collapse pit, the surface vertical displacement of the hanging wall and footwall was decreased by 56.8% and 55.6%, respectively, and the control effect of surface rock movement was remarkable.


Introduction
Underground non-coal mines are usually mined by caving method or empty field method [1,[8][9][10]25].This will inevitably form a certain scale of goaf in the stope.When the span of the goaf exceeds the critical caving span with the continuous mining [11,30,31], the roof rock mass of the goaf will cave.After the caving develops to the surface, it will lead to surface rock movement and large subsidence pits, which pose a serious threat to the surface environment and operators.In order to effectively predict the development trend of rock movement, Sidki-Rius et al. [32] used the method of field monitoring and numerical simulation analysis to enable the evaluation and the prediction of potential surface effects and improve the safety and environmental levels of the mining area.Woo et al. [34] developed a comprehensive database; the data clearly show that caving-induced surface deformations tend to be discontinuous and asymmetric due to large movements around the cave controlled by geologic structures, rock mass heterogeneity, and topographic effects.Ding et al. [4] proposed an analytical model to analyze the stability of the cylindrical caved space by employing the long-term strength of the surrounding rock mass, the in situ stress, and the impact of caved materials as inputs.The proposed model is valid for predicting the orientation and depth where rock failure occurs.The development of collapse and rock movement range depends on the break form and structural plane distribution of the surrounding rock [6,7,13,14].On the other hand, it depends on the lateral pressure support of caving and filling granular media.Engineering practice shows that the implementation of granular filling in the underground stope can effectively control the roof caving and surface rock movement [5,24,29,39],).Therefore, it is necessary to study the support mechanism of the granular media to the surrounding rock and achieve effective prediction and control of surface collapse and rock movement [22,23,30,31].
In recent years, researchers have carried out extensive research on granular lateral pressure.Kamiloglu et al. [15] developed an algorithm to obtain earth thrust coefficients depending on various wall dimensions and internal friction angles and to prepare graphs representing earth thrust coefficients and failure surface angles.Mirmoazen et al. [28] conducted a detailed numerical study to evaluate the lateral earth pressure acting on geosynthetic-reinforced retaining walls with an anisotropic granular backfill subjected to strip footing loadings.Kazempour S et al. [16] examined the compression behavior of aggregate-expanded polystyrene (EPS) granular composites using large-size oedometer tests measuring directly the coefficient of lateral earth pressure at rest (K0).Widisinghe et al. [33] presented vertical stress isobars for trenches filled with granular materials, developed from numerical modeling, and proposed in terms of dimensionless variables.These isobars can be used for determining vertical stresses at any point within the granular material contained in trenches and can become a simple and valuable tool.Chen et al. [2] used the limit equilibrium method and the whole analysis method of sliding wedge to provide the active earth pressure under the failure mechanisms of one to three slip surfaces.Xie et al. [35] presented the critical dimension of retained soils between the Coulomb failure plane and the planar slip surfaces influenced by rock faces, which could accurately predict the active earth pressure on rigid retaining walls built near rock faces.He et al. [12] proposed the backfilling of the granular rock into the collapse pit can prevent the surface strata movement.Kobyłka et al. [17] proposed that opening a discharge gate may result in a sudden increase in the lateral pressure with a simultaneous ramp down of vertical pressure.Li et al. [20] explored the influences of the compressive deformation of backfill materials on the strata movement and stress evolution in deep backfill mining and pointed out that the lower the compression ratio of backfill materials, the better the control of overlying strata.Li et al. [21] pointed out that the lateral pressure of confined granular material will be influenced by both the wall movement and the confined material width.Li et al. [19] pointed out that a reasonable particle size distribution can significantly improve stress characteristics, reduce crushing of particles in the samples, and increase the stiffness of the samples, so as to achieve better compaction effects.Le et al. [18] proposed a constitutive model that takes the real physical characteristics of granular material into account was proposed with variable deformation modulus, and the effects of the particle size of the granular backfill and the height and buried depth of mined-out stopes on surface subsidence have been clarified.Xu et al. [36] used the discrete element method (DEM) to study the lateral earth pressure change of granular materials under lateral strain-controlled cyclic loading, and a unique relationship is revealed between the deviator stress ratio and the deviator fabric of strong contacts on the micro-scale.These studies all have a common understanding the lateral pressure provided by granular media can play a certain supporting role and can be applied to the rock movement control in mining.
The waste rock produced by mining has the characteristics of non-uniformity.With the progress of underground mining, the granular media of the overburden is constantly moving, which leads to a more complex interaction between loose blocks, and cannot be displayed intuitively.At the same time, reasons such as particle size distribution, underground ore drawing, and ore body inclination will have an impact on the lateral pressure of the granular media [26].Through the analysis of a large number of literatures, there are few studies on the control of rock movement in underground mining based on the effect of granular lateral pressure, and the mechanism is less mentioned.In this paper, these reasons are considered; we use the self-developed granular lateral pressure experimental system for the first time to simulate and study the granular lateral pressure support mechanism under the synergistic effect of ore drawing and filling, which fills the research gap in this area.At the same time, combined with theoretical analysis, the calculation formula of granular lateral pressure and critical granular column height suitable for mine rock movement control is proposed for the first time.The theoretical value is in good agreement with the field-measured value, which further proves the reliability and applicability of the theoretical calculation formula.It can be applied to the rock movement control of steeply inclined ore body mining.This work helps researchers to better understand the mechanism of the granular media lateral pressure and has certain guiding significance for the effective control of ground subsidence and rock movement caused by underground mining.

Analysis of the Supporting Effect of the Granular Media on the Surrounding Rock
If the ore-rock contact surface of the collapsed pit is broken and there is no granular media support, the broken rock mass will be upright and fragmented along the ore-rock contact surface.When the granular media in the pit is high, the surrounding rock of the sidewall is stable, and the surrounding rock on the other side is intensified because of the lack of granular media support (Fig. 1).By analyzing the failure process of the surrounding rock on the sidewall of the collapse pit, when the caving granular media in the collapsed pit moves downward, the lateral support force exerted by the top granular media on the sidewall will unload.Under the combined influence of reasons such as stress, self-weight stress, and structural plane, cracks are formed along the structural plane toward the collapse pit.The sidewall rock will become unstable as the cracks expand, causing the surrounding rock to collapse or slip damage.Lateral caving and framing of the rock mass will lead to the continuous expansion of the subsidence pit.If the buckling block is supported by the lateral support of the granular media, the fractured block will be stable (Fig. 2).

Experimental Analysis of the Granular
Media Lateral Pressure

Experimental Model and Materials
Underground mining and collapse pit filling is a dynamic development process, which needs to be filled several times according to the flow of the granular media, but the numerical method can not accurately simulate this dynamic development process.Therefore, this paper combines similar physical experiments with theoretical analysis to clarify the rock movement control mechanism, which is used to guide on-site collapse pit treatment and rock movement control.In this paper, the granular media lateral pressure is abbreviated as "GMLP."The research adopts the lateral pressure experimental to study the variation law of the GLMP.The size of the model is l50 cm × 20 cm × 160 cm (length × width × height), the model similarity ratio is 1:100, and there are 16 CSF-1A sensor acquisition channels on both sides.The top of the experimental device is provided with a granular media inlet, and four granular media drawing ports are set at the bottom of the device, with a size of 3 cm × 3 cm.The data is displayed on the computer in two ways: the curve view and the data view (Fig. 3).The experiment is based on the particle size distribution of the waste rock powder in Xilinhot Fluorite Mine, and dolomite is used as the experimental granular media.The gradation of the experimental granular is shown in Table 1.
In this study, the selection of GLMP experimental degree was mainly based on the actual background of the Xilin Hot fluorite mine and the Gongchangling iron mine.The degree of the two mines was 80 ~ 90°.Therefore, the dip angle range of the experiment was 80 ~ 90°.The experiment selected three groups of ore body dip angles of 80°, 85°, and 90°.The single discharge of the granular media was about 1 kg.During the discharge, the released granular media should be backfilled into the model in time to ensure the height of the granular media did not change significantly.In the experiment, the side away from the outlet was the upper plate, and the side near the outlet was the lower plate.

Experimental Process
The steps of the side pressure experiment of granular media are shown in Fig. 4.
(1) Adjust the experimental equipment to 90°, open the data test device, check whether the 16 test channels are normal, and push the moving shaft rod connected to each sensor to the outer end to keep stability after confirmation.
(2) The bottom four outlets are blocked, and the configured bulk is slowly and evenly poured into the experimental device.A total of eight layers are filled, and the loading amount of each layer is about 36 ~ 40 kg.Since one sensor is responsible for the dispersion height of 20 cm, when the dispersion height reaches 20 cm, the data of the opposite side pressure is manually collected and saved until the whole device is filled with the particles.Eight times of data collection are carried out to obtain the value of the dispersion side pressure in the static state.height of the bulk accumulation does not change significantly.Then, manually collect the side pressure value of each channel until the side pressure value of each channel does not fluctuate, and obtain the active lateral pressure value of the granular media in the moving state.
(4) Reload the granular medium, and then open the ore drawing port for continuous ore drawing.The side pressure system is set to automatically collect data every 2 s.In the experiment, the passive pressure is applied to the particle medium in order to simulate the continuous deformation of the sidewall rock mass and automatically collect the passive pressure value under the movement of the particle medium.
(5) After the end of one inclination experiment, all the dispersion is released.The inclination of the equipment is adjusted, and experimental steps are repeated for the next set of experiments.
It should be noted that in the lateral pressure experiment of different angles, in order to ensure the effective comparability of the measured lateral pressure of the granular body, the overall loading height should be consistent, and the discharge times and overall discharge of the granular media should be consistent.

Analysis of Experimental Results
(1) Experimental Results of the GLMP.
The overall performance of the granular media side pressure was a monotonous increase, which was mainly divided into two stages: the fast growth stage and the slow growth stage.As the granular media height increased, the lateral pressure of the granular media increased rapidly at first.When the contact area between the granular media gradually became saturated and solid, the strong interaction between the granular media weakened, causing the lateral pressure to change gradually from a rapid increase to a slow increase, and gradually stabilized (Fig. 5).
The study found that with the increase of the dip angle, the pressure value of the footwall gradually decreased, and the average decrease rate was about 4.5%, and the critical depth was about 0.8 m.The pressure on the hanging wall gradually increased, and the average increase rate was about 7.9%; the depth was about 1.0 m.It showed the lateral pressure of the granular media on the footwall was more likely to be stable.With the change of the dip angle, the pressure value of the footwall was correlated negatively with the dip angle, and the pressure value of the hanging wall was correlated positively with the dip angle.The variation range of the GLMP on the hanging wall was higher than that of the footwall.The effect of the GLMP on the hanging wall was more prominent.
The GLMP curve under the condition of 90° inclination was shown in Fig. 6.With the release and backfilling of the granular media, the GLMP with a depth of 160 ~ 100 cm on the footwall decreased.When the amount of granular media released reached about 50 kg, the lateral pressure changed smoothly.The GLMP with a depth of 80 cm increased slowly and then changed steadily, indicating the granular media had met the critical depth.The GLMP with a depth of 40 cm and 60 cm gradually increased, indicating the granular media at this position would not move significantly with the release of the bottom granular media.The critical granular media column existed in the granular media above this position.The lateral pressure at the depth of 160 cm and 140 cm on the hanging wall gradually decreased.The variation trend of the lateral pressure at the depth of 120 cm and 100 cm could be divided into four processes.When the released granular media was 0 ~ 20 kg, the lateral pressure increased rapidly and reached a peak value.When the released granular media was 20 ~ 35 kg, the GLMP decreased and then rebounded.When the released granular media was 35 ~ 60 kg, the GLMP changed smoothly.When the released granular media was 60 ~ 85 kg, the GLMP increased slowly.It showed the loose range of the granular media was at the depth of 140 cm.The decrease of the GLMP after reaching the peak value might be related to the arching between the granular media.After the arch failed, the granular media moved down rapidly to fill the bottom space, resulting in a decrease of the GLMP.
The GLMP curve under the condition of 85° inclination was shown in Fig. 7. On the footwall, the GLMP decreased with a depth of 160 ~ 120 cm, and the average decrease rates were 14.17%, 10.87%, and 7.23%, respectively.Different from the 90° inclination angle, the GLMP with a depth of 100 cm began to increase, and the average increase rate reached 2.96%, showing the loose range was at the depth of 120 cm.On the hanging wall, the variation trend of the GLMP with the depth of 160 ~ 100 cm was the same as that of the 90° inclination angle.The average decrease rate of the GLMP with the depth of 160 cm and 140 cm was 6.54% and 4.07%, respectively.The average increase rates of the GLMP and the depth of 120 cm and 100 cm were 5.72% and 8.56%, respectively.
The GLMP curve under the condition of 80° inclination was shown in Fig. 8. On the footwall, the GLMP with a depth of 160 ~ 120 cm decreased, with an average decrease rate of 10.93%, 7.50%, and 4.01%, respectively.The average increase rate of the GLMP with a depth of 100 cm was 4.36%.The loose range was at the depth of 100 cm.The variation trend of the GLMP at the depth of 80-20 cm was the same as that of the 85° inclination angle.On the hanging wall, the variation trend of the GLMP with a depth of 160 ~ 120 cm was the same as that of the 85° inclination angle.The average increase rates of the GLMP were 6.96% and 6.47%, respectively.
Research shows that under the condition of keeping the height of the granular media constant, during releasing the granular media, only the GLMP near the ore drawing port reduce.The granular media near the ore drawing point moves from the hanging wall to the footwall, resulting in the loose range of the granular media on the hanging wall is higher than that on the footwall.If the height remains unchanged, the underground ore drawing will not affect the stability of the surrounding rock of the collapsed pit.With the granular media moving, the GLMP will increase outside the pressure decrease zone, and this part of the granular media will play a key role in restricting the nearsurface rock movement.The mass drawn (kg) The mass drawn (kg)

Construction of the Mechanical Model of the GLMP
According to the expansion research of the Janssen GLMP theory [3], the calculation formula of the GLMP is expressed as follows: where γ is density of the granular media, kN m −3 ; S is horizontal projected area of the granular media, m 2 ; α is inclination angle of the ore body, °; l is horizontal projected perimeter of the granular media, m; k w is GLMP coefficient; µ is friction coefficient between the granular media and the sidewall; µ = tanδ, δ is friction angle between the granular media and the sidewall, °; and h is depth of the granular media, m.According to formula (1), it can be seen the GLMP is affected by various reasons, and the mechanical analysis model is established based on the interaction between the granular media and the sidewall of the collapsed pit (Fig. 9).OA and BC are upper and lower boundaries of the wall; α is the inclination angle of the wall (ore body).OAC section is taken as the analysis object in the study, AC section is the slip surface (1) between the granular media, and β is the angle of the slip plane and the horizontal plane.DEFG at a certain depth is taken from the OAC section for mechanical analysis.The thickness is dy, the vertical load acting on the top is q, the bottom load is q + dq, and the lateral load acting on the wall is p w .The friction force between the wall and the granular media is f w , and the friction angle between the wall and the granular media is δ.The reaction force of the granular media on the slip surface is P r , and the friction angle in the granular media is φ.The friction angle between the granular media is φ, the friction force is f r , and the unit body weight is d w .
According to the relationship in Fig. 8, it can be obtained: (2) The friction between the wall and the granular media is expressed as follows: The frictional force between dispersions is expressed as follows: According to the balance of the force in the horizontal direction, the expression is as follows: Substituting formulas (2), (3), (8), and (9) into formula (10), the following expression is obtained by simplification: According to the balance of the force in the vertical direction, the expression is as follows: (10) The mass drawn (kg) The mass drawn (kg) Fig. 9 The mechanics model of the granular media lateral pressure Assuming the dispersion is not cohesive, the expression is as follows: where k w is the coefficient of the GLMP.
Substituting formulas (2), ( 3), ( 6), ( 7), ( 8), (9), and ( 13) into (12), omitting the second-order parameters, the following expression is obtained: Taking the center point Q of the slip surface between the granular media as the moment center, according to the moment balance condition, the expression is as follows: Substituting formulas ( 2), ( 3), ( 4), ( 5), ( 6), ( 7), (8), and (13) into formula (15), omitting the second-order parameters, the following expression is obtained: Combining formulas ( 14) and ( 16), the expression is as follows: Finally, the coefficient of the GLMP is expressed as follows: (12) In the Xilinhot Fluorite Mine, the friction angle between the granular media and the wall is 9°, and the friction angle in the granular media is 25°.The ore body dip angle varies from 80 to 90°, and the angle between the slip plane and the horizontal plane is β = 48°.We bring these parameters into Eq.( 18), and the GLMP coefficient under different dip angles can be solved.Where the GLMP coefficient of the 80° ore body dip angle is put into formula (1), the comparison between the theoretical value and the experimental value of the GLMP is shown in Fig. 10.
The calculated GLMP based on the theoretical is consistent with the experimental value basically, which verifies the reliability of the theoretically obtained GLMP.
Liu et al. [27] combined theoretical analysis with experiments and gave a formula for determining the height of the critical granular media column based on the GLMP: where H 3 is the height of the dense granular media column, m; H 2 is the total height of the granular media, m; H 4 is the height of the loose granular media column, m; P max is the maximum value of the GLMP when it reaches a steady state under the influence of ore drawing.
According to the research of ore drawing theory, the height of the loose body is about 2.5 times of the caving height of the overburden [24], and the calculation formula of the caving height of the overburden is as follows: where h b is the height of the ore house, m; h c is the height of the residual ore after the ore drawing, m; L is the thickness of the top pillar, m; ξ is the fragmentation coefficient.
Take According to the experimental results of the GLMP in Fig. 7, the GLMP increases significantly and stably with a depth of 1.2 m.Based on the experimental value, the theoretical formula (1) is revised.Taking l ⋅ h∕S as a dimensionless The GLMP at 1.2 m depth in the lateral pressure experiment with different inclination angles are put into formula (22) and analyzed by Matlab software to get the correction coefficients R 1 = 0.5, R 2 = 8.15, R 3 = − 1.95.The improved calculation formula of the GLMP is expressed as follows: Substituting parameters into formulas ( 18), (23), and (19), the comparison between the theoretical value and the fieldmeasured value of the critical granular media column height in two mines is shown in Fig. 11.
It can be seen from Fig. 10, the theoretical value of the critical granular media column height obtained according to the improved calculation formula is close to the field-measured value relatively.The improved GLMP calculation formula can be better applied to engineering.
According to the theoretical research of the critical granular media column [37,38], the boundary of the surface collapse area extends downward to the boundary of the ore body.The granular media column above this position is the critical granular media column (Fig. 12).The relationship between the slump angle and the critical depth is as follows: where H is the mining depth, m; β is the slump angle, °; β 0 is the rock displacement angle, °; α is the ore body dip angle, °.Combining formulas (19) and (24), the maximum allowable unfilled height of the collapse pit under the effective control of collapse and rock movement is obtained as follows: ( 22) The GLMP proposed by Chen [3] is mainly aimed at the pressure support effect of filling granular body on the floor pillar of the steep inclined orebody and does not reflect the important role of sidewall rock mass control.On this basis, through theoretical analysis, this paper gives the calculation formula of the GLMP coefficient.At the same time, combined with the experimental results of the GLMP, three correction coefficients R 1 , R 2 , and R 3 are introduced.Finally, the improved calculation formula of the GLMP is determined, and the calculation formula of the critical granular media column is further determined.Through the research, this formula can be used to fill mine collapse pits, which can achieve effective control of surface collapse and rock movement.

Rock Movement Control Mechanism
Based on the experimental results of the GLMP, the granular media near the drawing point first loosens, and the compaction density of the loose body gradually decreases, resulting in the GLMP on the surrounding rock decreases.After the end of the ore drawing, the flow trend of the loose body near the ore drawing port from the hanging wall to the footwall slows down.Finally, the hanging wall loose body slows down and solidifies, and the final movement trace of the granular media is parallel to the wall of the surrounding rock (Fig. 13).When the backfilling granular media provides enough lateral support force, the surrounding rock of the collapse pit will no longer be fragmented, and the surface collapse and rock movement expansion will be controlled.
The GLMP largely depends on the density of the granular media, and the density of the granular media depends on the granular media height.With the increase of the granular media height, the granular media actively exerts a lateral pressure σ 1 on the sidewall.At the same time, it is affected by the deformation of surrounding rocks, the granular media squeeze each other, resulting in a decrease in porosity and an increase in the number of contact points.The strong interaction between particles will bear the lateral pressure σ 2 exerted by the sidewall rock mass (Fig. 14).The loose area at the bottom is mainly affected by the lateral pressure σ 1 , and the upper dense area is jointly affected by the lateral pressure σ1 and σ 2 , which provides support for the effective action of the upper critical granular media column.
Comprehensive analysis shows the granular media in the collapse pit are mainly divided into three action regions: the loose granular media column, the compacted granular media column, and the critical granular media column 15).The change of the lateral pressure of the loose body in the action area of the loose granular media column will not affect the surface collapse and rock movement.In the action area of the compacted granular media column, the granular media always keeps a high-density compacted state.The GLMP will increase until it tends to be stable, providing stress support for the upper critical granular media column.The critical granular media column is the most important area to control surface collapse and rock movement and it provides lateral support.The lateral support force it provides can keep the rock mass stable within a certain height above the critical granular media column, to control the surface collapse and rock movement.

Engineering Background
The Xilinhot Fluorite Mine belongs to China Iron and Steel Group Mining Company and is located in Xilinhot City, Inner Mongolia Autonomous Region, northwest China.The wall rock of ore-rock contact zone is mainly altered diorite, with 0.6 ~ 1.2 m alteration, and the stability is poor.The wall rock near the ore is mainly biotite plagioclase granite.The fluorite grade is between 20.05 and 96.20%, with an average of 63.21%.The orebody intendancy ranges from 100 to 105°, the dip angle ranges from 80°to 90°, the horizontal thickness ranges from 4 to 12 m, average thickness is 7 m, and the elevation is 480 m.The thickness and occurrence of the ore body are stable, and the mine adopts shaft development and high-end wall sublevel caving mining.The stage height is 40 m, the sublevel height is 20 m, and the caving step is 1.6 m.The mining roadways are arranged along the vein, and the mining route is close to the footwall of the ore body.The ore recovery rate is 82.01%, and the dilution rate is 17.89%.The stope structure is shown in Fig. 16.

Surface Collapse and Rock Movement Control Methods
Taking Xilinhot Fluorite Mine as an example, because of the influence of the short-hole shrinkage method, an obvious collapse pit was formed on the surface.The dip angle of the collapse pit is 86°.According to Fig. 11a, the critical granular media column height of the collapse pit is 35 m.The mining depth of this mine is 225 m, the slump angle is 83°, and the rock movement angle is 65°.Substituting the relevant parameters into formulas ( 24) and ( 25), the maximum allowable unfilled height of the collapse pit is calculated as 2.5 m.Subsequently, the collapse pit was fully filled with waste rock extracted from the shaft until the filling height requirements were met.The surface was covered with soil, and the soil height was 2.5 m.In order to verify the effect of surface rock movement control, the change characteristics of the surface subsidence before and after the collapse pit filling were monitored.The layout of the monitoring points was shown in Fig. 17.The monitoring points were arranged perpendicular to the strike of the ore body.Eight monitoring points were arranged on the hanging wall and footwall of the collapse pit, with four monitoring points on each side, and the average distance between the monitoring points was 4 m.The total station was used for monitoring, and the monitoring period was once every half month, the monitoring period was about 3 months.
The monitoring results are shown in Fig. 18.As the distance between the monitoring point and the collapse pit decreased, the vertical displacement showed an overall increasing trend, and the vertical displacement of the hanging wall was higher than that of the footwall.When the collapse pit was not filled, the maximum vertical displacement of monitoring points 4# and 5# near the collapse pit was 19.2 cm and 19.8 cm, respectively.After the collapse pit was filled, the maximum vertical displacement of monitoring points 4# and 5# near the collapse pit was 8.3 cm and 8.8 cm, respectively.The vertical displacement decreased by 56.8% and 55.6, respectively.
After nearly a year of observation, a fracture line appeared on the hanging wall of the collapsed pit (Fig. 19), which was caused by the slight dislocation of the granular media during the settling process, and the surface did not

Conclusions
(1) Through the experimental study of the GLMP, the static GLMP shows two stages of rapid growth and slow growth with the increase of the depth.The lateral pressure on the hanging wall is smaller than the footwall.The influence of the GLMP on the hanging wall rock mass is more prominent.When the granular media filling height remains unchanged, the ore drawing only causes the GLMP near the drawing port to decrease, and the decrease range is less than or approximately equal to the critical depth value.The GLMP outside the lowering zone is not affected by the ore drawing.
(2) The research gives an improved calculation formula for the GLMP.By comparing with the experimental and field-measured data, the theoretical calculation formula can be well applied to calculate and analyze the GLMP and the height of the critical granular media column under different ore body conditions.
(3) Under the influence of ore drawing, the granular media in the collapse pit are mainly divided into three action zones: loose granular media column, compacted granular media column, and critical granular media column.The change of the GLMP in the action area of the loose granular media column will not affect the surface collapse and rock movement.The lateral support force provided by the granular media in the critical granular media column action area is the key reason affecting the rock movement.If the critical granular media column height requirements are met, the surface collapse and rock movement will be controlled.(4) Through surface rock movement monitoring, as the distance from the sidewall of the collapse pit decreases, the vertical displacement shows an overall increasing trend.The vertical displacement on the hanging wall is higher.The height of the critical granular media column in the collapse pit is 35 m, and the maximum allowable unfilled height of the surface is 2.5 m, which can guide the effective filling of the collapse pit.
(5) Through field practice, after effective filling of the collapse pit, a small fracture line appeared on the hanging wall because of the slight dislocation of the granular media in the collapse pit, and the overall stability was good.Compared with the collapsed pit without filling, the vertical displacements of the hanging wall and the footwall were reduced by 56.8% and 55.6%, respectively.This method realizes the effective control of the surface collapse and rock movement of the Xilinhot Fluorite Mine.

Fig. 5
Fig. 5 GLMP curves affected by the granular height under different inclination angles

Fig. 6
Fig.6 The GLMP curve under the condition of 90° inclination

Fig. 7
Fig.7 The GLMP curve under the condition of 85° inclination

Fig. 8
Fig.8 The GLMP curve under the condition of 80° inclination Xilinhot Fluorite Mine and Gongchangling Iron Mine as examples.In the Xilinhot Fluorite Mine, h b = 26 m, h c = 1.5 m, ξ = 1.5, L = 14 m, and γ = 3.18 × 10 3 kg/m 3 .In the Gongchangling Iron Mine, h b = 45 m, h c = 2 m, ξ = 1.5, L = 15 m, and γ = 4.1 × 10 3 kg/m 3 .Substitute the relevant parameters into formula (21), we can get the drop heights of the overburden in two mines are 17.5 m and 35.5 m, respectively, and determine the heights of the loose body are 43.75 m and 88.75 m, respectively.

Fig. 11 5 Fig. 12
Fig. 11 Comparison of theoretical and measured values of the critical granular medium column

Fig. 16 Fig. 17
Fig.16 The stope structure of the short-hole shrinkage method

Fig. 18
Fig.18 Monitoring results of surface rock movement (a) Before filling the collapse pit (b) After filling the collapse pit

Fig. 19
Fig. 19 Comparison of collapse pit before and after filling