Design of Key Parameters for Strip–Filling Structures Using Cemented Gangue in Goaf—A Case Study

Abstract

Large–scale underground coal mining is bound to cause serious surface subsidence problems. However, conventional filling and mining methods have problems such as high cost and process difficulty. In order to achieve the purpose of high efficiency and low cost, this paper proposes using the technology of CGSG. To achieve the effective control of overburden strata movement and ground surface settlement using cemented gangue strip filling in the goaf (CGSG), this paper studies the design principles and methods of the key parameters of the strip–filling structure including the strength, compressed deformation characteristics, and sizes. Based on the analysis of the structures and movement characteristics of the overburden strata above the coal seam, the mechanical relationship between the strip–filling structure and the overburden strata was established. Formulas for calculating the parameters of the strip–filling structure were derived. Guided by the obtained index parameters, the material ratios and mechanical experiments of the filling body were designed. The research results demonstrated that the strengths of the cemented gangue filling body at different ages should be greater than the compressive load of the strata roof movement on the filling body during the same period; under the compression of the maximum load, the ultimate compressive deformation of the filling body should be less than the ultimate subsidence deflection of the basic roof strata. The width of the strip–filling structure was inversely proportional to its ultimate strength, while the width of the non–filling area was greatly affected by the length of the rock beam formed after the basic roof strata fractured. The research results were applied in the No. 7402 experimental strip–filling workface in Zhaizhen coal mine, China. Reasonable parameters of the cemented gangue strip–filling structure were designed. The field application results demonstrated that, after using the technology of CGSG, there was no obvious pressure appearance when the working face was mined. The maximum sinking value of the ground surface was only 30 mm after the mining of the working face was completed; at the same time, the filling cost was about one–third less than the complete–filling technology in the goaf.

1. Introduction

The underground mining of coal can cause ground surface subsidence after a large number of coal seams have been mined. This can seriously damage the land and water resources in the mining area. At the same time, a large amount of gangue accumulates on the surface, which can spread to cultivated land, pollute the environment, and endanger the health and safety of people. Gangue is a kind of solid waste generated along with coal production in coal mines. How to deal with gangue is a hot issue in mining. Gangue backfill mining technology can dispose of the gangue directly into the goaf. This is one of the most reasonable ways to solve gangue problems from the perspective of the coordinated development of mining and environment protection [1,2,3,4,5]. The development of goaf backfill mining technology has been researched for nearly a hundred years. Scholars from many countries, such as Australia, China, Poland and the United States, have carried out extensive research in various aspects such as filling materials, filling theory and technology in the goaf. Many valuable research results have been obtained. The technologies of hydraulic filling, solid gangue filling and cemented high–water material filling have achieved good results in field engineering applications [6,7,8,9,10,11,12,13,14]. Study by Malashkevych, D et al. For the first time, a mechanism for the formation of operational ash content and energy value of coal has been revealed when combining the processes of drifting operations to prepare reserves from new extraction pillars with associated stope operations into a new selective mining technology with waste rock accumulation in the mined–out area. Additionally, through further research results, new space–planning solutions were designed to achieve the best maximum storage of rock in the underground area of the mine [15,16]. The study by Ghiasi, V et al. predicted the ground surface settlement in the tunneling of a single circular tunnel with simultaneous changes in the mechanical properties of the soil and the geometrical properties of the tunnel section using artificial neural network analyses [17]. The study by Azarafza, M. et al. presented a procedure for establishing a deep–learning–based predictive model (DNN); the model was able to accurately predict the geotechnical indices [18]. The above findings have important implications for this study.

The technology of complete filling the goaf with gangue still has some obvious disadvantages, including the large demand for waste gangue, complicated filling processes, high filling cost, and a large effect on production. These disadvantages limit its widespread application in mines. In response to the above problems, some mining scholars have proposed technical solutions for strip filling in the goaf. The technical principle is shown in Figure 1. Strip–filling bodies are constructed in the goaf along the mining advancement direction of the working face. These bodies are used to dispose of the gangue waste and also support the overburden strata. The supporting function of the strip–filling structure can control the large–scale sinking deformation of overburden strata, thereby achieving the purpose of controlling mining subsidence and protecting the mining area environment.

The technology of CGSG can achieve the same technical effect as complete–filling technology in the goaf. However, it has obvious advantages, including reducing the filling cost, simplifying the filling process and improving the filling and mining efficiency of the working face. These have great practical significance for improving the extraction rate of coal resources, protecting the land and water resources of the mining area, extending the service life of the mine and developing the local economy. Therefore, the technology of CGSG has great potential in mining.

Some scholars have carried out in–depth research on the technology of strip filling, including the technical principle, the action mechanism and supporting effect between a strip–filling body and roof strata, the parameter design and the stability criterion of strip–filling structures. These studies have made some important research conclusions [19,20,21,22]. However, as a new type of filling technology, the current research on strip–filling technology is still in its infancy and has not yet formed a sound theoretical system and technical content.

The design of the key parameters of the strip–filling structure is essential for the successful application of this technology. The key parameters include the size, the bearing strength and the ultimate compression amount of the strip–filling body. This paper takes a specific engineering case in Xinwen mining area in China as the research object and focuses on optimizing the design of the key parameters of the filling body in the strip–filling technology with the comprehensive methods of theoretical calculation, numerical simulation and experiment. The study aims to ensure the long–term stability of the strip–filling structure and overburden strata roof in the goaf. The research aims to establish a reasonable and effective design method for strip–filling parameters.

2. Engineering Background

Zhaizhen coal mine is located in the Xinwen mining area in Eastern China, and the annual output of the mine is 1.2 million tons. Due to the high number of mining levels and the many rock roadways that need to be excavated, a large amount of gangue is produced in this mine. At the same time, the ground surface damage caused by underground mining is serious, and the deformation of the surface building is larger. To deal with the gangue produced in the mine, while also controlling the overlying strata stability and reducing the ground subsidence, the technology of CGSG was proposed. After theoretical and technical feasibility studies, it was decided to test this technology in the 7402 working face of the No. 7 mining district in Zhaizhen coal mine.

2.1. Introduction of 7402 Strip–Filling Face

The 7402 strip–filling test working face was located in the upper part of the No. 7 mining district of the −400 m mining level in Zhaizhen coal mine. The mining depth was approximately 375.3 m to 400.3 m. The main coal seam mined from this working face was No. 4, which has a thickness of 2.17 m and a dip angle of approximately 8°. The 7402 working face length was 160 m, and its strike length was 655 m. The recoverable reserves of coal in the area of the working face were about 64,000 tons. The strike longwall coal mining method and comprehensive mechanized mining technology were used. The goaf management was controlled with the method of strip–filling using cemented gangue. Figure 2 presents the planar layout view of the 7402 working face and its adjacent roadways. An auxiliary track alley is arranged as a special filling roadway, the main filling equipment such as the gangue mixer is arranged in the roadway, the filling material is sent into the goaf area of the working face through the filling and conveying pipeline, the upper end of the filling roadway and the 7402 transport lane intersection are set up with gangue silos to store the gangue produced by each excavation face underground, the gangue is unloaded to the feeding port of the crusher through the feeder and the discharge conveyor belt, the crusher crushes the gangue to particles with a particle size of no more than 35 mm, and the gangue powder and granules are transported through the feeding belt. The machine is sent into the mixer, water is added to the mixer along with cement, fly ash and additives according to the ratio to fully mix into a cemented body, and then the chute is used to slip to the transfer pump and transport it to the goaf area through the conveyor pipe for filling.

2.2. Analysis of the Surrounding Strata Structures

According to the geological analysis of the No. 7 mining district in Zhaizhen coal mine, the roof strata of the upper No. 4 coal seam are mainly fine sandstone, medium sandstone and siltstone. The above rocks belong to moderately hard rocks with a hardness coefficient of approximately 4–5. Among the above main rock strata, there are several weak rock layers, including mudstone, sandy mudstone and argillaceous siltstone, with a hardness coefficient of about 2–3. According to the strata structure analysis of the mining district, the roof strata composition of upper No. 4 coal seam is complicated and the lithology of the surrounding rocks is quite different. The overlying strata are composed of several layers of hard and soft rock stratum of various thicknesses. In particular, there are some layers of thick sandstone and siltstone strata with large thickness (>10 m) in the overlying strata above the No. 4 coal seam. Based on the high difference in thickness and lithology, as the No. 4 coal seam is mined, the movement of the overlying strata would not be coordinated with each other. This movement should demonstrate a certain range of a group movement phenomenon.

According to the comprehensive analysis of the strata structure and the mechanical properties, the entire overlying strata above the No. 4 coal seam should belong to the medium–strength category. According to the geological comprehensive histogram of the No. 7 mining district, the main strata structure and their mechanical characteristics above the coal seam are listed in

Table 1. The overlying strata structures and main rock parameters above coal seam in No. 7 mining district.

Order	Lithology	Thickness (m)	Depth (m)	Bulk Density (kN/m3)	Compressive Strength (MPa)	Elastic Modulus (GPa)
1	Topsoil	170	170	17.0	/	/
2	Red sandstone	83	253	24.6	56.2	6.9
3	Clay rock	34	287	17.5	10.2	0.65
4	Mudstone	10.4	297.4	20.5	18.5	1.82
5	Medium–fine sandstone	6.6	304	24.0	48.6	6.0
6	Fine sandstone	1.89	305.89	25.2	58.5	7.5
7	Medium sandstone	7.54	313.43	24.1	50.8	6.8
8	Sandy mudstone	6.55	319.98	21.1	20.6	2.2
9	Silt–fine stone	18.3	338.28	26.0	64.5	8.2
10	Medium–fine sandstone	4.2	342.48	24.0	48.6	6.0
11	Fine sandstone	3.15	345.63	25.2	58.5	7.5
12	Sandy mudstone	7.8	353.43	21.1	20.6	2.2
13	Siltstone	6.5	359.93	23.4	34.8	5.5
14	Muddy siltstone	4.35	364.28	22.5	21.4	2.3
15	Fine sandstone	8.46	372.74	25.2	58.5	7.0
16	Medium sandstone	6.35	379.09	24.1	50.8	6.8
17	Fine sandstone	15.6	396.69	25.2	58.5	7.5
18	Muddy siltstone	1.9	398.59	22.5	21.4	2.3
19	Mudstone	0.5	399.09	20.5	18.5	1.82
20	No. 4 coal seam	2.17	401.26	18.0	15.8	1.02

.

3. Theoretical Foundations for the Key Parameters of the Strip–Filling Body

3.1. Design of Strip–Filling Body’s Strength

The main principle of the CGSG technology is to construct a certain width of strip–filling bodies in the goaf. The overlying rock strata are all well supported by these strip–filling bodies and the key stratum layer or the basic roof stratum will not fracture. Thus, the purpose of reducing the pressure appearance of the working face and the ground subsidence can be achieved. According to the above principle of the CGSG technology, the strength of the strip–filling body is a key parameter to ensure the success of this technology.

The strength design of the strip–filling body is closely related to the overlying strata structure and its motion characteristics at different time periods. There are different requirements of the strength of the strip–filling body in different stages of the roof strata’s movement. First, as the coal seam is mined, the immediate roof will break and fall down in a short period of time. Thus, the strip–filling body should have sufficient early strength to support the whole weight of the falling immediate roof. Second, as the mining area continues to expand, the basic roof starts to move and causes a significant pressure appearance on the working surface. At this stage, the strip–filling body should have sufficient medium–term strength to control the basic roof’s movement without fracture damage. Finally, when the mining area of the working face reaches the full mining effect, the strip–filling body should have sufficient late strength to control the bending settlement of the entire range of rock strata from the coal seam to the ground surface. The early, middle and late strengths of the strip–filling body are closely related to the loads caused by the movement of the rock strata. Meanwhile, the range of the moving strata is affected by the mining area of the working face at different time periods. Based on the above analysis, before designing the strength of the strip–filling body, it is necessary to analyze the influence range of the overlying strata in different stages after the coal seam mining. Then, the next step is to determine the structure and movement characteristics of the affected overlying strata range at a certain stage, and calculate the specific loads that the strip–filling body should support.

3.1.1. Discrimination of Composite Strata Structure

Extensive studies have demonstrated that the overlying strata move in strata groups, or units, after coal seam mining because of the differences in stratification characteristics of coal and rock seams in the strata. In each strata group, the role of each coal and rock stratum layer in the rock strata movement is different. The movement of the relatively thick and hard rock stratum layer controls the movement of the entire strata group. Therefore, this rock layer is the bearing body and skeleton within the strata group. This type of rock stratum is defined as the key layer and is generally located at the bottom of each strata group. The relatively weak and thin rock strata in each strata group are generally attached to the key layer, which is controlled by the movement of the key layer, and together with the key layer, the strata form a simultaneous movement strata group [23,24]. When the key layer moves, the sinking deformation of the upper soft rock layer is coordinated with it. The motion parameters (fracture step, sinking curvature, etc.) of the key layer are the motion parameters of all the rock layers of the entire strata group.

In a certain rock strata group, finding the location of the key layer and determining its fractured step and breaking shape is of great significance for studying the overall motion law of the overlying strata. The strata groups’ discrimination was determined by the strength factors of each stratum, including the lithology, thickness and hardness coefficient. An upper stratum with a low strength factor will move with a lower rock that has a high strength factor. The maximum curvature of the stratum subsidence could be used to determine if the adjacent strata move independently or as a strata group [25,26].

When ${\rho }_{\max 1}\ge {\rho }_{\max 2}$, these two strata combine into a strata group;

When ${\rho }_{\max 1}<{\rho }_{\max 2}$, these two strata move separately, and they belong to two different strata groups.

Here, ${\rho }_{\max 1}$ is the maximum subsidence curvature of the upper stratum and ${\rho }_{\max 2}$ is the maximum subsidence curvature of the lower stratum.

According to material mechanics theory, the maximum subsidence curvature of a fixed stratum beam can be described by the following formula:

$${\rho }_{\mathit{max}}=\frac{\gamma L^2}{2E{m}^2}$$

(1)

where L is the hanging span of the stratum; E is the elastic modulus of the stratum; and m is the stratum thickness.

According to the strata structure in the No. 7 mining district and the mechanical parameters provided in Table 1, the aforementioned criteria that can be used to determine the overlying key stratum layers above the No. 4 coal seam include:

• The key layer I: the No. 17 fine sandstone layer with 15.6 m thickness and 58.5 MPa compressive strength.
• The key layer II: the No. 9 powder fine sandstone layer with 18.3 m thickness and 64.5 MPa compressive strength.
• The key layer III: the No. 2 red sandstone layer with 83 m thickness and 56.2 MPa compressive strength.

The location of the key rock layers in the overlying strata and the strata range each layer controls are shown in Figure 3.

According to the analysis of overlying strata structure in No. 7 mining district, the No. 18 and No. 19 rock layers are 2.4 m thick and belong to the immediate roof. This strata group will collapse and fill the goaf as the mining space gradually increases. Strata group I (No. 10 to No. 16 rock strata) is 56.41 m thick and controlled by the key layer I belonging to the basic roof. As the mining area increases, when the main key layer I starts to move, it will drive the overlying soft rock layers in the strata group to move at the same time. Then, a separation space will appear between it and strata group II controlled by the main key layer II. Strata group II (No. 3 to No. 8 rock strata) is 56.41 m thick and controlled by the key layer II. As the mining area continues to increase to reach the extreme motion step of this key layer, it will sink. Then, the separation space between it and strata group I will gradually be compacted. Strata group III controlled by the key layer III includes all the strata above the key layer up to the ground surface. Its total thickness is 253 m. If this key layer breaks, the ground surface will experience significant subsidence.

3.1.2. The Movement Characteristic and Load Analysis of Overlying Strata

After the coal seam was mined, the subsidence movement of the overlying strata gradually developed from the bottom to the top. Due to the different structure of the overlying strata group, the subsidence movement characteristic of the roof strata was different. Therefore, the load acting on the strip–filling body in the goaf was different at the different stages of the subsidence of the overlying strata.

(1)

The early–to–medium load analysis

The working face started to advance from the open–off cut, and the mining area gradually increased with the working face moving forward. In the mining process of the working face, the strip–filling bodies were constructed in the goaf, and the profile view along the mining direction of the working surface is shown in Figure 4. When the advancement length of working face L was less than the fractured step length l1 (that is, the ultimate overhang length when the rock beam broke first) of the key layer I, This period can be considered as the early stage of the overlying strata movement. During this stage, because the key layer I (basic roof) did not break, it and the entire strata group I could maintain the transmission of force in the layer direction. The dead weight load of the entire strata group was mainly supported by the solid coal walls on both sides. At this time, the load acting on the strip–filling body in the goaf mainly came from the immediate roof (including the No. 18 and No. 19 rock layers in Table 1).

The two rock layers contained in the immediate roof were generally soft, so they were easy to collapse with the mining of the working face. The mining height of the 7402 working face was 2.17 m, and the thickness of the immediate roof was 2.4 m. After the immediate roof collapsed, the bulk expansion coefficient of the gangue was about 1.3–1.5; thus, the stacked gangue could not fully fill the goaf, and it was not possible to contact the roof plate in the unfilled area. This means the stacked gangue in the goaf could not provide a supporting capacity for the overlying strata. Therefore, the total weight of the immediate roof above the strip–filling body was supported by the filling body.

Based on the above analysis, the load $q_1$ required to be carried by the strip–filling body was mainly the dead weight load of the upper immediate roof in the early stage of the overlying strata movement. The load can be calculated by the following formula.

$$q_1={\displaystyle \sum }_{i=18}^{19}{\gamma }_i\cdot h_i$$

(2)

where ${\gamma }_i$ is the bulk density of the rock stratum, kN/m³; h_i is the rock stratum thickness, m.

As the working surface advances further, when the fractured step length l1 of the key layer I is less than the advancement length of working face L, which is less than the fractured step length l2 of the key layer II, this period can be considered as the mid–term stage of the overlying strata movement, as shown in Figure 4. In this stage, the mining area is sufficient for the adequate movement of main key layer I. If there is not enough supporting force below, the key layer I will sink and break. This will cause the large–scale settlement deformation of the roof strata. Therefore, when the mining distance of the working surface exceeds the fractured step length l1 of the main key layer I, the strip–filling body should reach the sufficient medium strength in this stage. The strip–filling body should be able to provide the overlying strata group I sufficient supporting force to avoid the fracture of the basic roof from reaching its ultimate sinking deflection.

When the strata group I controlled by the key layer I moves, the suspended length of strata group II controlled by the key layer II has not reached its ultimate fractured step length. Therefore, strata group II does not undergo large–scale settlement deformation. A normal separation space will appear between strata group II and strata group I. The existence of this space interrupts the mechanical continuity and transfer of force between the two strata groups.

According to the valid district theory and considering the greatest safety factor, the load $q_2$ required to be carried by the strip–filling body in the mid–term stage is mainly the sum of the dead weight load of the upper immediate roof $q_1$ and the dead weight load of strata group I.

The load can be calculated by the following formula.

$$q_2=q_1+{\displaystyle \sum }_{i=10}^{17}{\gamma }_i\cdot h_i$$

(3)

(2)

The late load analysis

As the working surface advances further, when the mining distance exceeds the fractured step lengths of key layer II and key layer III, these two key layers start to move fully. This period can be considered as the late stage of the overlying strata movement, as shown in Figure 5. In this stage, the overburden strata group above the basic roof (strata group I) will gradually undergo a bending and sinking movement and the normal separation spaces between each strata group will gradually be compacted. The overburden loads originally supported by the solid coal walls on both sides will gradually transfer to the strip–filling bodies in the goaf. In this stage, the load $q_3$ required to be carried by the strip–filling body is the dead weight loads of all the rock strata above. In order to effectively control the ground subsidence, the strip–filling body should reach sufficient late strength in this stage, thus preventing the surface collapse caused by the fracture of the overlying key strata.

The load $q_3$ can be calculated by the following formula.

$$q_3=q_1+q_2+{\displaystyle \sum }_{i=1}^9{\gamma }_i\cdot h_i$$

(4)

3.1.3. The Strength Design Requirements of the Strip–Filling Body

The main components of the cemented gangue filling body are gangue waste, cement, fly ash, water, and other admixtures. The main factors affecting the strength of the filling body include the type and amount of cement, particle grading of the gangue waste, and the stirring and curing time.

In the material composition of the cemented gangue filling body, cement is the main acting binder. Its solidification shows obvious staging during the curing process. The solidification strengths of the cement occur at 8 h, at 3 days, and at 7 days with the ultimate strength after 28 days [27,28]. Therefore, the hardening process of the cemented gangue filling body is also affected by the hardening process of the cement. After the strip–filling body is constructed in the goaf, its strength development is also staged. It exhibits gradual strengthening characteristics in different periods, with grading characteristics of early strength, medium–term strength and late ultimate strength.

According to the above analysis, the load applied on the strip–filling body is phased, and the solidification hardening process of the filling body is also phased, and the two are related. Based on the control mechanism of the strip–filling structure on overlying strata movement, the corresponding relationship between the strength of the gangue–cemented strip–filling body and the overburden load can be established.

• The early strength: ${\sigma }_E\ge q_1$

• The mid–term strength: ${\sigma }_M>q_2$

• The late strength: ${\sigma }_L>q_3$

3.2. Design of Ultimate Compression Amount of Strip–Filling Structure

In the process of strip filling, to avoid large settlement deformation on the surface, the basic roof and above key rock layers are not allowed to break. For a certain overlying stratum, there is a limit sinking deflection w₀ during its movement, and when the sinking deflection is exceeded, the rock stratum will break. It is assumed that the strip–filling body is always connected to the roof during its hardening process; therefore, the amount of compression of the filling body is consistent with the overburden strata subsidence amount. Therefore, the allowed ultimate compressive deformation of the strip–filling body itself must be less than the ultimate sinking deflection of the basic roof under the compression of maximum load, which is u_max < w₀. Otherwise, if u_max > w₀, the strip–filling body cannot effectively support the overburden roof, and the roof will be broken and damaged. Thus, determining the ultimate subsidence deflection of the basic roof plays an important role in guiding the material ratio of the filling body, determining its ultimate compressive capacity and mechanical properties.

For the basic roof, the ultimate subsidence deflection is close to the maximum when the suspension length reaches its first fracture step length. At this point, if the strip–filling body can effectively support the rock layer so that it does not break, then the basic roof will remain intact during its subsequent periodic movement.

The first fracture step length of the basic roof is relatively small compared to the length of the working surface, so its maximum subsidence deflection can be calculated by the theory of the rock beam. Before the basic roof beam breaks, it can be considered that its two ends are in a fixed state, forming an embedded beam structure, as shown in Figure 6.

According to the calculation method of the sinking deflection of the beam in material mechanics, for the embedded beam structure in which two ends are fixed, the maximum subsidence deflection when bending and sinking can be expressed by the following formula:

$$w_0=\frac{\gamma l^4}{32E{m}^2}$$

(5)

where $\gamma$ is the bulk weight of the stratum; l is the rock beam’s ultimate fracture step length; E is the elastic modulus; and m is the thickness of a certain stratum.

3.3. Width Design of the Strip–Filling Body and Unfilled Zone

In the technology of strip filling in the goaf, since the basic roof and above overlying strata are not allowed to break and fracture, a reasonable width of the filling body and the unfilled area must be designed to ensure a sufficient support strength to the overlying strata. The width of the filling body and the unfilled area is affected by multiple factors, such as the ultimate fracture step length of the rock beam, the overburden load, and the supporting capacity of the filling body.

(1)

Design based on the rock beam’s ultimate fracture step length

According to the theory of rock strata movement and mining pressure control, when the suspension length of the basic roof reaches its ultimate fracture step length, the rock beam will break and lose its bearing capacity. So, in theory, based on the premise that the filling body strength and the filling width are sufficient, the rock beam of the overlying basic roof can be stable and will not fracture as long as the width b of the unfilled area is less than the ultimate fracture step length l1 of the overlying basic roof, as shown in Figure 7. l1 can be calculated from plate theory [29]. This can effectively control the settlement movement of other upper key stratum layers and the subsidence of the ground surface.

(2)

Design based on the overburden load and supporting capacity of the filling body

According to the previous analysis of the load supported by the filling body, in the later stage of the strata movement, the entire weight of the overlying strata above the goaf will be supported by the strip–filling body [30]. Therefore, the sum of the bearing capacities of all strip fillings within the goaf must be capable of supporting the entire weight of all overlying strata within the same area. That is, the weight of the overburden strata in the unfilled area will be supported by the adjacent strip–filling bodies. The mechanical relationship can be expressed by the following formula:

$${\displaystyle \sum _{i=1}^n{\gamma }_i{h}_i}\cdot (a+b)\le \sigma \cdot a$$

(6)

where n is the total number of rock strata above the coal seam to the ground surface; a is the width of one strip–filling body; b is the width of one unfilled area; and $\sigma$ is the ultimate supporting strength of the strip–filling body.

4. Production and Mechanical Testing of the Filling Body

The main components of the gangue–cemented filling body include gangue waste, cement, fly ash, water and other additives. Different combinations of these components will result in different mechanical properties. The basic principle of the combination for the gangue–cemented filling body is to obtain the best mechanical strength at a reasonable cost. According to the analysis of the strip–filling body’s key parameters, successful strip–filling has clear requirements for the strength and ultimate compression of the filling body. To obtain a reasonable filling body, an experimental study on the material ratio and mechanical properties of the components was carried out.

4.1. Production

Due to the diversity of the composition of the filling materials, based on the previous research results of the proportion of the gangue–cemented filling materials, an orthogonal experimental method was used to design this experimental scheme. In the material composition of this experiment, the coarse aggregate was −35 mm gangue waste, and the fine aggregate was river sand. The cement consisted of two parts, C40 grade ordinary Portland cement and fast and hard sulphoaluminate cement. The fast and hard sulphoaluminate cement accounted for 10% of the total cement quality, which can improve the solidification time and early strength of the filling body. In the experimental scheme, three levels and four factors were designed for optimization analysis. The selected four factors were fly ash content, gangue aggregate content, mass concentration and the early strength agent dosage. Where the cement content was taken as the base unit 1, the amount of fly ash and gangue aggregate could be expressed by the corresponding ratio. According to the basic material ratio, three experimental levels including different material ratios were designed, as shown in

Table 2. The testing factors and levels.

Factors	Fly Ash Content	Mass Concentration (%)	Gangue Aggregate Content	Early Strength Agent Dosage (%)
Code	A	B	C	D
Upper level (1)	0.4	80	7	1.2
Middle level (2)	0.5	82	8	1.4
Lower level (3)	0.6	84	9	1.6

.

4.2. Mechanical Testing

After producing the standard testing specimen by using the filling materials of the respective ratios, the conservation of different ages was carried out. According to the solidification age characteristics of the cement and the time period required for the filling body to reach the strength of each stage, it was determined that the conservation age of the test specimen in this experiment w 1 d, 4 d, 14 d and 28 d, respectively. When the conservation time of the sample reached its age, the sample was subjected to uniaxial compression experiments. The experiment was carried out on the MTS815.03 electro–hydraulic servo rock experimental system of the State Key Laboratory of Mining Disaster Prevention and Control in Shandong University of Science and Technology, China. The uniaxial compressive strength results of the testing samples obtained in the test are shown in

Table 3. The testing scheme and results.

Designed Matrix					Uniaxial Compressive Strength (MPa)
Schemes	A	B	C	D	8 h	4 d	14 d	28 d
1	1	1	1	1	0.65	5.34	10.55	13.88
2	1	2	2	2	0.58	5.24	9.88	10.20
3	1	3	3	3	0.49	4.30	7.22	8.58
4	2	1	2	2	0.78	6.32	11.21	12.85
5	2	2	3	1	0.46	4.05	6.33	9.38
6	2	3	1	3	0.47	4.26	5.08	8.98
7	3	1	3	2	0.41	4.84	7.205	10.66
8	3	2	1	3	0.46	3.66	5.79	8.02
9	3	3	2	1	0.23	2.06	4.56	5.58

. The uniaxial compressive strength of each ratio at different times is shown in Figure 8.

According to the testing results, the strength of the filling body at 8 h was generally less than 1 MPa, mainly concentrated at 0.4–0.6 MPa. After solidification for about 4 days, the strength of the filling body increased rapidly and could reach about 4–6 MPa. After that, the strength of the filling body still increased slowly, and the final strength could reach about 10 MPa. In addition, among the four factors of material ratio, the gangue aggregate content was the main factor affecting the middle strength and ultimate strength of the filling body, because it directly affected the proportion of cemented materials in the whole material. Fly ash is a kind of weakly cemented material, and its amount played a role in the strength of the filling body after 14 days. In addition, fly ash had a greater influence on the fluidity and pumpability of the filling slurry. The early strength agent and mass concentration played a major role in controlling the solidification time and slump of the filling body, and, therefore, had a great influence on the early strength of the filling body.

5. Design Results and Discussion

5.1. Strength of the Strip–Filling Body

(1)

Calculation of the overlying strata load

According to the analysis of movement characteristics of the roof strata and the action load on the strip–filling body, the movement of the roof strata is phased and, therefore, the load applied to the strip–filling body at different stages is different. Formulas (2)–(4) in Section 3.1.2 can be used to calculate the loads at different stages.

① Putting the mechanical parameters of the immediate roof (in Table 1) in the load calculation Formula (2), the load applied to the strip–filling body in the early movement stage of the overburden strata was 0.053 MPa.

② Putting the mechanical parameters of the strata from the coal seam to strata group I (in Table 1) in the load calculation Formula (3), the load applied to the strip–filling body in the middle movement stage of the overburden strata was 1.45 MPa.

③ Putting the mechanical parameters of all the strata above the coal seam (in Table 1) into the load calculation Formula (4), the load applied to the strip–filling body in the late movement stage of the overburden strata was 8.19 MPa.

(2)

The hardening time calculation of the strip–filling body

In the early stages of the roof movement, the first filling body was constructed when the work surface started to advance from the open–off cut. When the advancing length reached the ultimate fractured step length of the immediate roof, the first strip–filling body began to effectively support the entire weight of the overlying immediate roof within its width. Therefore, during this time period, the filling body should reach its early strength ${\sigma }_E$. The early strength ${\sigma }_E$ should meet the requirements ${\sigma }_E>q_1=0.05\mathrm{MPa}$. According to the on–site production arrangement, the 7402 working surface advanced slowly due to the impact of the filling process. The advanced mining speed of the working face was about 3.6 m/d. According to theoretical calculations and on–site observation analysis, the ultimate fractured step length of the immediate roof was approximately 12 m. Therefore, the time required for the strip filling–body to reach its early strength was about 12/3.6 = 3.3 days.

The mid–stage of the overlying strata movement began when the main key rock layer I reached its ultimate fracture step length. In this stage, the weight load of the entire strata group I gradually acted on the strip–filling body. The middle strength ${\sigma }_M$ of the strip–filling body should meet the requirements ${\sigma }_M>q_2=1.45\mathrm{MPa}$. The movement of strata group I was controlled by the key rock layer I which was 15.6 m thick. According to theoretical calculations and on–site observation analysis of the 7402 working face, the ultimate fractured step length of the key rock layer I was approximately 88 m. Considering the construction time and hardening time of the strip–filling body, combined with the field practice experience, the strip–filling body generally started to be constructed when the working surface’s advancing length reached half of the ultimate fracture step length of the key rock layer I (i.e., there were about 2–3 strip–filling bodies under the rock beam for each ultimate step length). Based on the analysis of the working face’s advancing speed of 3.6 m/d, the minimum time required for the strip–filling body to reach its mid–term strength was approximately (88/3.6)/2 = 12.2 days.

According to the above analysis, the late strength of the strip–filling body in the late stage of the roof movement should not be less than 8.19 MPa, and the minimum time required was about 25.3 days.

Table 4. The requirements for strengths and hardening times of the strip–filling body.

Strip–Filling Body	Early Stage	Middle Stage	Late Stage
Strength (MPa)	${\sigma }_E$ $\ge$ 0.05	${\sigma }_M$ > 1.45	${\sigma }_L$ > 8.19
Minimum time (days)	3.3	12.2	25.3

shows the requirements for strengths and hardening times of the strip–filling body at different stages.

(3)

Discussion

According to the mechanical testing results of the filling samples, the final strength of the filling samples in schemes 8 and 9 (<8.19 MPa) could not meet the design requirements in Table 2. The results of schemes 1–7 all met the hardening time and strength index requirements of each stage in Table 2.

In addition, based on the analysis of the strip–filling body’s width and the mechanical relationship Formula (6) in Section 3.3, we found that the width of the strip–filling body was inversely proportional to its strength index. That is, the lower the ultimate strength of the filling body, the larger the width that needs to be designed. The increase in the width of the filling body will inevitably increase the filling cost and the difficulty. Therefore, to ensure a reasonable width of the strip–filling body, schemes 1 and 4 that have larger ultimate strength can be considered for further comparison.

The basic principle of the combination of a gangue–cemented filling body is to obtain the best mechanical strength at a reasonable cost. Although the final ultimate strength of scheme 1 was slightly larger than scheme 4, the medium–term strength of scheme 1 was small; that is, the hardening speed of scheme 1 was not as fast as that of scheme 4. At the same time, in terms of filling cost, scheme 1 reduced the proportion of gangue aggregate, which inevitably increased the amount of cement and other materials, so the overall filling cost was greater for scheme 1.

Based on the comprehensive analysis, scheme 4 was chosen to be the optimal material ratio for the strip–filling body of the 7402 filling face in Zhaizhen coal mine. The filling body made by scheme 4 had a hardening strength of no less than 0.78 MPa at 8 h and a hardening strength of 6.32 MPa at 4 days, which met the requirements for the early strength of the filling body in Table 2. Its hardening strength at 14 days and 28 days could, respectively, reach 11.21 MPa and 12.85 MPa, which met the requirements for the middle and late strength of the filling body in Table 2. The test results in this paper were obtained according to the comprehensive analysis of the special geological conditions of Zhaizhen coal mine, and are only applicable to the filling test surface of 7402 of Zhaizhen coal mine. The research method in this paper required the recalculation of the specific geological conditions for different working faces, and the construction efficiency and construction cost needed to be considered when the strength was satisfied.

5.2. Compression Amount of the Strip–Filling Body

According to the analysis in Section 3.2, putting the corresponding parameters of the basic stratum (key layer I) into the calculated Formula (5) showed that the maximum subsidence deflection w₀ before it broke was 23 mm. Thus, the ultimate compression amount of the strip–filling body in the goaf should be smaller than 23 mm to control the basic stratum from fracturing.

The full stress–strain curve after 28 days under uniaxial compression of the filling body made by scheme 4 is shown in Figure 9.

The maximum compressive strain of the testing sample in the experiment under the action of the maximum compressive stress of 12.85 MPa was about 0.0078, and the compressive strain was about 0.0042 under the compressive stress of 8.19 MPa (the maximum load of the overlying rock acting on the filling body). The mining height of the 7402 filling working face was about 2.17 m. Assume that the strip–filling body was always in contact with the roof stratum during the movement of the roof. Then, in the late stage of the roof movement, the compressive deformation amount of the strip–filling body was $2.17\times {10}^3\times 0.0042=9.114 \mathrm{mm}$ when the maximum applying load (8.19 MPa) was reached. The compressive deformation amount of the strip–filling body could reach $2.17\times {10}^3\times 0.0078=16.926 \mathrm{mm}$ when the applied load reached the ultimate strength of the filling body (12.85 MPa).

According to the above calculation, the maximum compressive deformation of the filling body formed by the ratio of scheme 4 was only 9.1 mm when it was subjected to the maximum load in the late stage of the roof. Even if the applied load reached the ultimate strength of the filling body, its maximum compressive deformation was only 16.926 mm. These two values were lower than the ultimate sinking deflection w₀ = 26.0 mm of the basic stratum roof. Therefore, it can be determined that the compression deformation of the filling body made by scheme 4 satisfied the requirements.

5.3. Size Parameters of the Strip–Filling Body

In the analysis in Section 3.3, to ensure that the basic roof and above rock strata did not fracture, the width b of the unfilled area in the goaf was required to be smaller than the fractured step length l1 of the overlying basic roof beam. In Section 5.1, the ultimate fractured step length of the key rock layer I was calculated as approximately 88 m. Thus, b < 88 m. At the same time, to achieve the static mechanical balance in the vertical direction, it was necessary to satisfy the mechanical relationship (5), which was ${\displaystyle \sum _{i=1}^n{\gamma }_i{h}_i}\cdot (a+b)\le \sigma \cdot a$. This formula demonstrates that the width size of the filling body was inversely proportional to its ultimate strength. The higher the ultimate strength of the filling body, the smaller the required width size, and, conversely, the larger the required width size.

In the later stage of the roof movement, the overburden load required to be supported by the strip–filling body was about 8.19 MPa, that is ${\displaystyle \sum _{i=1}^n{\gamma }_i{h}_i}=8.19\mathrm{MPa}$. According to the mechanical testing results of the filling samples, the ultimate strength of the sample in scheme 4 was about 12.85 MPa. Thus, Formula (6) could be converted into 8.19(a + b) 12.85a. Then, the relationship between the width of the strip–filling body and the width of the unfilled area could be obtained as: $a\ge 1.76b$.

To achieve the best filling effect, while reducing the filling cost as much as possible, the width of the unfilled area was designed to be b = 20 m, and the width of the strip–filling body was designed to be a = 40 m. This designed result also considered the engineering geological conditions and the filling process requirements of the 7402 filling face in Zhaizhen coal mine. The design scheme could basically guarantee that there were 1–2 complete strip–filling bodies to effectively support each basic roof beam.

5.4. Engineering Application Effect

To evaluate the control effect of strip–filling technology on overburden strata movement and ground surface subsidence, strata behavior monitoring and ground subsidence observation were carried out. These works were carried out at the same time as the filling process of the 7402 working face. The monitoring results were also compared with the working face where the goaf was completely unfilled.

(1)

Analysis of the strata behavior

Figure 10 shows the curves of the support’s working resistance with the increase in the working face’s advancing distance. The data in the upper curve were collected from the working face where the goaf was completely unfilled; the data in the lower curve were obtained from the 7402 strip–filling face. The two observed working faces were located in the same mining district and had similar geological conditions.

The upper curve shows that the working resistance of the hydraulic support was periodically increased when the roof strata were allowed to fall freely in the goaf. This demonstrates that the basic roof was periodically fractured, and when it occurred, a larger pressure load was applied to the support, causing a pressure appearance. The curve also shows that the normal working resistance of the support was about 30–32 MPa. When the working face’s advancing distance was about 86 m, the basic roof fractured for the first time, and the maximum resistance of the support was up to 52 MPa. This demonstrates a high degree of stress concentration. As the working face continued to advance, an obvious pressure appearance caused by basic roof fracture occurred three times. The advancement distance of the working face was about 30–32 m between the two adjacent pressure appearances and the working resistance of the hydraulic support was about 40–46 MPa at each facture of the basic roof.

The lower curve shows that the working resistance of the hydraulic support had no obvious change within 200 m of the working surface advancement. It did not have a significant resistance increase zone like the upper curve. This was mainly due to the effective control of the overburden roof by the strip–filling body in the goaf of the 7402 working face, so that the basic roof and overburden strata did not fracture. In the 200 m range of working face advancement, the working resistance of most hydraulic support was always maintained at around 25–28 MPa, and the maximum did not exceed 32 MPa. Therefore, there was no obvious pressure appearance during the advancement of the working face.

The above analysis shows that the strip–filling technology in the goaf could effectively control the overlying strata movement similarly to complete–filling technology. Based on the design method of the key parameters of the strip–filling body proposed in this paper, reasonable parameters were designed for the 7402 filling face of Zhaizhen coal mine, which had a good effect on the control of overlying strata movement.

(2)

Analysis of the ground subsidence

Ground surface subsidence observation is the most intuitive and effective method to judge the control effect of strip–filling technology on ground surface deformation. For the working face where the roof strata were allowed to fall freely in its goaf, the maximum sinking value of the ground surface was about 700–750 mm caused by the mining activity and the ground surface deformation was very serious.

After the completion of the mining of the 7402 filling face, the observation base station was established in the corresponding position of the ground surface. The obtained surface subsidence curve is shown in Figure 11. The curve in Figure 11 demonstrates that the maximum sinking value of the ground surface in the mining–affected area was less than 30 mm, and the average sinking amount was about 14 mm. This means that the designed strip–filling parameters of the 7402 working face were reasonable, and the ground surface deformation was well controlled.

The observation and analysis of underground strata behavior and surface subsidence of the 7402 working face which used the strip–filling technology in the goaf showed that the strip–filling body could well support the key rock layer without fracture, thus controlling the movement of the entire strata and reducing the subsidence deformation of the ground surface. The strip–filling parameters designed for the 7402 working face are reasonable and the design method for a strip–filling body proposed in this paper is also correct.

Based on the same controlling effect on the strata movement and ground subsidence, the strip–filling technology used in Zhaizhen coal mine can reduce the filling cost by about one–third compared to the complete–filling technology in the goaf. This can create considerable economic benefits for mining companies.

6. Conclusions

There are six conclusions. First, in the mining technology of cemented gangue strip filling in the goaf, the key parameters of the strip–filling body ensure the filling effect. These parameters include the strengths at different stages, compressed deformation characteristics, and the sizes of the filling body.

Second, the overlying strata movement is staged after the coal seam is mined; therefore, the load acting on the filling body in the goaf is different during different movement stages. Meanwhile, due to the influence of the cementing materials in the composition of the filling body, the hardening time and strength growth of the cemented gangue strip–filling body also have staging characteristics. Thus, in the same stage, the filling body’s cementing strength should be greater than the load applied on it.

Third, the basic principle of the strip–filling body to control the overburden roof is to ensure that the key rock layer (basic roof) does not fracture. Therefore, under the maximum compressive load, the ultimate allowed compressive deformation of the strip–filling body should be smaller than the ultimate sinking deflection of the key rock layer.

Fourth, the width of the non–filled zone is greatly affected by the fracture step length of the basic roof beam. To ensure that the basic roof beam does not fracture, the width of the non–filling zone should be smaller than the ultimate fracture step length of the basic roof. Meanwhile, at least one or two strip–filling bodies should exist under a basic roof beam with the length of an ultimate fracture step. The width of the strip–filling body is inversely proportional to its ultimate strength. According to the engineering application experience, a suitable ratio of the width of the strip–filling body to the width of the non–filling zone is 1.8:1 to 2:1.

Fifth, according to the above analysis, the late strength of the strip–filling body in the late stage of the roof movement should not be less than 8.19 MPa, and the minimum time required is about 25.3 days. Based on the comprehensive analysis, scheme 4 was chosen to be the optimal material ratio for the strip–filling body of 7402 filling face in Zhaizhen coal mine. The filling body made by scheme 4 has a hardening strength of not less than 0.78 MPa at 8 h, and a hardening strength of 6.32 MPa at 4 days. Its hardening strength at 14 days and 28 days can, respectively, reach 11.21 MPa and 12.85 MPa.

Sixth, the designed method for the key parameters of the strip–filling technology was successfully used in the 7402 working face of Zhaizhen coal mine. The effect of the strip–filling technology was verified in practice. There was no obvious pressure appearance during the advancement of the working face. The amount of ground surface subsidence was reduced from the original 700–750 mm to 30 mm. The filling cost was reduced by about one–third.