Experimental Investigation on Mechanical Properties of In Situ Cemented Paste Backfill Containing Coal Gangue and Fly Ash

Cemented paste backfill containing coal gangue and fly ash (CGFACPB) is an emerging backfill technique for coal mines that allows environmentally hazardous coal gangue and fly ash to be reused in the underground goaf. Meanwhile, CGFACPB can provide an efficient ground support and reduce the surface subsidence. Due to the difference of consolidation environment between the laboratory and the field, the mechanical properties of the cemented paste backfill vary significantly. In this paper, the core specimens were collected from an underground coal mine where the CGFACPB was used for coal mining, and the mechanical properties of the collected specimens were investigated. ,e cores were obtained from the underground coal mine, and then the standard cylinders or discs were prepared in laboratory.,e uniaxial compressive strength (UCS), Young’s modulus, and Poisson’s ratio were determined by the compression tests, and the tensile strength was achieved by the Brazilian test. ,en the internal friction angle and cohesion were calculated using the improved Mohr–Coulomb strength criterion. ,e results showed the development of UCS can be divided into four stages, and the final long-term stable value was about 5.1MPa.,e development of Young’s modulus had similar trend. Young’s modulus had a range from 550MPa to 750MPa and the mean value of 675MPa. Poisson’s ratio gradually increased with the underground curing duration and eventually approached the stable value of 0.18. ,e failure type of compression samples was mainly single-sided shear failure.,e development of tensile strength can be divided into two stages, and the stable value of the tensile strength was about 1.05MPa. ,e development of cohesion can be divided into four stages, and the stable value was about 1.75MPa.,e stable value of the internal friction angle was about 25°.,is study can provide significant references for not only the long-term stability evaluation of CGFACPB in the field but also the design of optimal recipe of the cemented paste backfill (CPB).


Introduction
Cemented paste backfill containing coal gangue and fly ash (CGFACPB) is an emerging backfill technique in the coal mines in China. e cemented paste backfill can not only dispose of solid waste (mainly coal gangue and fly ash) but also significantly improve the recovery ratio of coal resources and reduce the surface subsidence [1][2][3]. In the deep underground coal mines in China, CGFACPB is used to replace the residual strip coal pillar; however, the in situ supporting performance of CGFACPB has been a major concern [1,2]. e consolidation process of CGFACPB occurs in the complicated underground curing and stress environment. Under the high overlying strata stress, CGFACPB material could collapse or experience an excessive deformation, which could not only affect the mining operations but also cause the serious damages to surface villages [3][4][5][6]. CPB is a widely used backfilling method in mining that can significantly improve stope stability in the metal mines or reduce the surface subsidence in the underground coal mines [1,7]. In general, CPB materials are produced by mixing solid waste (e.g., coal gangue and tailings), a single or composite hydraulic binder (e.g., Portland cement or slag cement), and water. e CPB materials have a high slump slurry (generally 18-25 cm) [1][2][3]6] and are transported into the underground stope by the pipeline system. e main properties of the CPB material include low binder content (3-7% based on dry weight) and a low uniaxial compressive strength (UCS), i.e., in the range of 0.5 MPa to 5 MPa [4,6,7].
At present, most of studies have been focused on metal mines (e.g., gold, iron, and copper), in which the primary materials were tailing, hydraulic binder, and water, defined as TCPB (tailing cemented paste backfill). Yilmaz et al. [8] investigated the effect of curing time of TCPB with different types of binder and content on the one-dimensional consolidation characteristics using an improved laboratory apparatus. Ghirian and Fall [9] studied the behaviors in the thermal, hydraulic, mechanical, and chemical processes in 1, 3, and 7 days. Cihangir et al. [10] discussed the effect of activator type, concentration, and slag composition on the strength and stabilities of high-sulphide tailing CPB with alkali-activated slag as the binder. Yilmaz et al. [11,12] analyzed the effect of specimen size, curing time, and stress conditions on the strength and hydromechanical, geotechnical, and geochemical properties using a new laboratory apparatus. Fall et al. [13,14] investigated the stress-strain behaviors under the uniaxial compression and the conventional triaxial compression at the humidity of approximately 80% and the temperature of 23 ± 2°in the laboratory. Sivakugan et al. [15][16][17][18] conducted a stress-strain test under unconfined and confined pressure in the laboratory. Aldhafeeri et al. [19][20][21] proposed a relationship between the increase of reactivity and the mechanical damage level of CPB by analyzing the durability and environmental behavior of CPB in the chemical reactivity. Ercikdi et al. [22] prepared the sulphide tailings CPB using the granulated marble waste as an additive and the waste bricks as the replacement and additive to ordinary Portland cement. ey showed the using of marble waste samples as an additive could significantly enhance the short-and long-term mechanical strength. Koohestani et al. [23,24] conducted an experiment on the influence of maple-wood sawdust addition to the mechanical and microstructural properties. Koupouli et al. [25] performed an experimental study on the shear strength of CPB-CPB and CPB-rock wall interfaces. In addition, the fictional behaviors of these interfaces were assessed under the short curing durations using a direct shear apparatus. Based on previous studies, the understanding of engineering properties of TCPB has been significantly improved. As an emerging backfill technique in coal mine, CGFACPB has been studied in fewer studies. Wu et al. [26][27][28] evaluated the thermal, hydraulic, and mechanical performance of CGFACPB using ultrasonic testing. Yin et al. [29] investigated the effect of solid components on the rheological and mechanical properties of CGFACPB. e main focus of the above studies is the properties of laboratory-prepared CPB samples using the conventional moulds. However, some studies have indicated that even the CPB was prepared using the identical recipe under the same curing conditions and the UCS of field-prepared samples was usually 2-4 times higher than that of the laboratoryprepared samples [30][31][32][33]. e difference in the UCS of CPB might be due to the difference in the curing and stress conditions in the field and in laboratory. e conditions in the field improved the strength development rate and the ultimate strength. Due to the significant difference in geological mining conditions, backfilling process, and materials between coal mines and metal mines, the aim of this paper is to investigate the mechanical properties of in situ coring CGFACPB in the laboratory. e UCS, Young's modulus, and Poisson's ratio were determined by the uniaxial compressive tests, and the tensile strength was obtained by the Brazilian splitting test. Based on an improved Mohr-Coulomb strength criterion, the internal friction angle and cohesion were calculated. In addition, some of the damage types of CGFACPB were analyzed.

Materials and Methods
e core samples were collected from Daizhuang coal mine, located in Jining, Shandong Province, China. In this coal mine, approximately 80% of coal resources are buried under the villages. e thickness of coal seam is about 2.7 m, and the depth of overlying strata is about 500 m. e traditional longwall caved methods can introduce the severe surface subsidence and thus cannot be used to mine the coal seams. erefore, a longwall strip mining method was adopted in the coal mine. Currently, a total of 49 residual strip coal pillars exist in the coal mine with the total coal reserves of up to 10 million tons. In order to recycle the residual strip coal pillars, backfilling mining technique was introduced to the coal mine. e first field trial region was selected in the 2300 mining district, and the coal pillar between 2302 and 2303 panel was the targeted object, as shown in Figure 1. e dimensions of coal pillar were 110 m × 960 m (width × length). e CGFACPB recipe was determined by a material proportioning test in the laboratory. e ratio of cement : fly ash : coal gangue was 1 : 4 : 6, and the solid concentration was 74%. Detailed information on the recipe of CGFACPB can be found in our previous study [34].

Coring Process and Samples Preparation in the Field.
Due to the prominent difference in the mining and backfilling operations between metal mines and coal mines, there is a significant discrepancy in the backfilling process. For Daizhuang coal mine, the longwall mining method was used. erefore, there was a serious interference between coal cutting process and backfilling process. CGFACPB material was gradually backfilled into the goaf region as the working face advanced by about 4-5 m. e segregating operation was conducted to separate the working region and the backfilling region. Usually, a backfilling work was conducted for 3 days every time. e dimensions of the backfilled volume were 4 meters of width, 100 meters of length (longwall working face), and 3 meters of height (working face height). After 28 days in the laboratory, CGFACPB material reached the design strength [1][2][3]34]. At least 10 backfilling processes were required, and as a result, a depth of 40 meters borehole needed to be drilled to obtain the cores. e drilling process was carried out by the TXJ-1600 geological drilling rig (Luheng Instrument, Jining, China). e coring position was about 1.2 meters above the floor. Two drilling boreholes #1 and #2 were located in the Site I and Site II areas, and their locations are shown in Figure 2. During the coring process, the coring depth and sample numbers were recorded carefully. Plastic wrap was used to seal the samples to prevent the evaporation of water and the damage of sample.
ereafter, the samples were transported to the laboratory and processed to obtain the standard samples indoors (Φ50 mm × 100 mm compression test; Φ50 mm × 25 mm tensile test). e testing was conducted immediately. Core samples are shown in Figure 3.

Testing Methods.
Testing was performed in the MTS815.03 rock servo experiment system. e system was controlled by a computer, and the data were collected and recorded automatically. ere were three sets of independent servo instruments, which can control axial compression, confining pressure, and pore pressure. e servo valve can respond to the frequency of up to 290 Hz, and stretch sensors can work accurately at the high temperature (200°C) and oil pressure (140 MPa). e stress-strain curves before and after damage can be measured accurately. For the compression test, the displacement control was adopted, the prepeak loading speed was set to 0.1 mm/s, and the postpeak loading speed was set to 0.2 mm/s. e Brazilian disc testing method, proposed by International Society for Rock Mechanics, was used to measure the tensile of CGFACPB samples. In order to reduce the influence of the interface friction on the strength of the sample, the rigid ingot head was used. e MTS815.03 rock servo experiment system, the compression (UCS) test system, and tensile test system are shown in Figures 4 and 5, respectively.

Compression Test Results.
In this study, 232 standard samples were obtained from two drillings. Because the backfill was very fragile, fewer standard samples were obtained. On one hand, the CGFACPB material had a lower strength and lower coring ratio. As a result, it was very difficult to acquire long intact cores. On the other hand, damages occurred during the transportation and laboratory processing. ere were 121 cores for drilling #1 (Figure 2 Site I) and 111 cores for drilling #2 (Figure 2 Site II). From boreholes #1 and #2, there were 58 and 51 cores used for the compression test, respectively. e UCS, Young's modulus, and Poisson's ratio were obtained by the compression testing. Due to the large discreteness of the testing results, the mean values of samples with the same curing time in the underground were used to evaluate the testing results. Figure 6 illustrates the development of UCS with the underground curing time. e underground backfilling process can be used to interpret the evolution of UCS with the underground curing time and the roof stress environment, as shown in Figure 7. Combined with the process of filling mining, we interpret the USC change process of the filling body.
As we all know, the UCS of CPB obtained from the uniaxial compressive tests is the most important parameter to evaluate the property or recipe value of CPB. From the Advances in Civil Engineering fitting curve of UCS with the underground curing time, it can be observed that the evolution of UCS can be divided into four stages: (a) Rapid development stage I: in this stage, the consolidation of CGFACP changed from the slurry to the solid, and the initial strength value continuously increased from 0 MPa up to self-standing strength (generally, the curing time is 8 hours in the laboratory because the hydraulic supports move forward with the advancement of working face). In the working face, the consolidation conditions of CGFACP mainly included curing condition and stress condition. e curing condition consisted of temperature and humidity of the working face. e stress condition was dependent on the roof strata loading. Generally the curing condition of the first block backfill was different from that of other blocks because one side of the backfill was the working face and the other side was backfill. e first block backfill   was under the cover of face supports and the back backfill; thus the roof stress loading generally was little and often negligible according to the field experience. e value of UCS within three days was increased from 0 MPa to 2.72 MPa in drilling #1 and to 2.60 MPa in drilling #2, indicating that the growth velocity was rapid, which can be observed from the fitting curve. (b) Slow development stage II: when the second block backfill was gradually backfilled into the goaf region, the first block had been cured for three days in the underground. In this stage, the consolidation environment of the first block had changed not only the curing environment but also the stress environment. e curing temperature and humidity of the first block had significant change, and the roof stress was gradually increased. e UCS value was increased from 2. Young's modulus is another important index, which reflects the resistance of the material to elastic deformation. e larger Young's modulus indicates a greater stiffness of the material. From the fitting curve in Figure 8, the development trend of Young's modulus was similar to that of the UCS. us, the development of Young's modulus can also be divided into four stages. In the first stage, i.e., the rapid development stage, the CGFACPB slurry was changed from fluidity to solidity in the underground curing environment. In the second stage, i.e., the medium development stage, Young's modulus value was increased from the minimum of 475 MPa up to 610 MPa in drilling #1 and from 175 MPa up to 525 MPa in drilling #2. e third stage, i.e., slow development stage, lasted for 6 days. From Figure 8, there is a large discreteness in the value of Young's modulus in both drilling #1 and drilling #2. In the fourth stage, i.e., long-term development stage, the relatively stable trend can be observed from the fitting curve. Young's modulus was in the range of 550 MPa to 750 MPa. e value of Young's modulus in the fitting curve was 675 MPa.
Poisson's ratio is an elastic parameter to characterize the lateral deformation capacity of materials. Poisson's ratio reflects the ratio between the deformation magnitudes of the material in two different directions. Larger Poisson's ratio indicates that the transverse deformation is larger than the longitudinal deformation. Figure 9 shows the trend of Poisson's ratio with the underground curing time. Within the first twelve days, Poisson's ratio was gradually increasing in both drilling #1 and drilling #2. ereafter, the change of Poisson's ratio was not obvious. Poisson's ratio was between 0.17 and 0.19 in drilling #1 and between 0.19 and 0.28 in drilling #2. Poisson's ratio was gradually increased in the first 12 days and stable at 0.18.
Generally, the rock has three failure types under the unconfined or uniaxial compression loading, including X-conjugate oblique shear failure, single slope shear failure, and splitting failure [35]. CGFACPB material was a lowgrade concrete material, which was also called similar rock material. Some typical failures of different samples are shown in Figure 10. From the figure, the failure of the samples was mainly single-sided shear failure.

Tensile Test Results.
In this study, 63 blocks in drilling #1 and 60 blocks in drilling #2 were used in the tensile test. e Brazilian method was used to test the tensile strength of the samples. Figure 11 shows the development of the tensile strength with the underground curing time.
e investigation on the tensile strength of the sample can provide significant reference to not only the stability of cavern project but also the slicing backfill mining of thick coal seam. From Figure 11, the development of the tensile strength can be approximately divided into two stages. In the first stage, i.e., the development stage within the first twelve days, thè   e testing results of the tensile strength showed that CGFACPB material had better resistance to shear deformation. Some photographs of the typical damage are shown in Figure 12.

Cohesion and Internal Friction Angle.
In this study, in order to calculate the cohesion and the internal friction angle of CGFACPB, Mohr-Coulomb strength criterion was used [36,37]. e equation is as follows: where τ is the shear stress, σ is the normal stress, c is the cohesion, and ϕ is the internal friction angle. e Mohr-Coulomb strength criterion can be expressed by the Mohr limit circle, as shown in Figure 13.
σ and τ can be expressed by the principal stress σ 1 and σ 3 . e equation is as follows: where θ is the broken angle of rock and σ 3 are the principal stresses. Substituting equation (2) into equation (1), the Mohr-Coulomb strength criterion can be expressed as When σ 3 � 0, σ 1 is the σ of UCS, then the equation can be written as follows: erefore, the equation of the Mohr-Coulomb strength criterion in the σ 3 − σ 1 coordinate system is as follows: where σ c is the UCS.
Substituting tan φ � f and equation (2) to equation (1), the equation can be rewritten as follows: Taking the derivative of this equation to θ, the maximum of τ − fσ can be obtained as follows: If the value of equation (7) is less than c, the rock would not break; otherwise, the rock would break. If τ − fσ mas � c, equation (7) can be written as e intensity curve of the Mohr-Coulomb strength criterion in the σ 3 − σ 1 coordinate system is shown in Figure 14.
ereafter, σ can be expressed as follows:   Advances in Civil Engineering e minimum value of σ 1 can be used to determine when the rock is going to break in the maximum limit equilibrium state. If σ > 0 and any θ value, the following equation is obtained from equation (2): Since Since ����� � f 2 + 1 > 0, in the case σ > 0, the following equation can be obtained: e equation of the Mohr-Coulomb strength criterion in the σ 3 − σ 1 coordinate system can be expressed as follows: erefore, when σ 1 � 0, σ 3 � σ t <0, β of the straight line PL can be gained in the following equation: tgβ � 2σ t /σ c . erefore, the Mohr-Coulomb strength criterion in the σ 3 − σ 1 coordinate system can be determined by the value of UCS and the tensile strength indirectly. e equations of the internal friction angle and the cohesion can be expressed by the value of UCS and the tensile strength, as shown in the following equations: Based on the testing results of UCS and tensile strength, the value of the cohesion and the internal friction angle with  Figures 15 and 16, respectively. e cohesion and the internal friction angle were two important parameters to reflect the resistant capability to the shear strength of rock mass and soil mass. e development of cohesion of CGFACPB material can be divided into four stages. In the first stage, i.e., rapid development stage, within the first three days, the cohesion value was increased from 0 MPa to 0.94 MPa in drilling #1 and from 0 MPa to 1.08 MPa in drilling #2. In the second stage, i.e., medium development stage, from the third day to the ninth day, the cohesion value was increased from 1.08 MPa to 1.23 MPa in drilling #1 and from 0.94 MPa to 1.63 MPa in drilling #2. e development velocity in the second stage was slightly lower than that of the first stage. In the third stage, i.e., slow development stage, there was a certain discreteness in the value. In the fourth stage, i.e., gradual or long-term development stage, although the values had a large discreteness, a stable value can be observed from the fitting curve. e approached stable value from the fitting value was about 1.75 MPa.    e internal friction angle refers to the angle between the positive stress and the internal friction stress, which is formed when the rock approaches the limit equilibrium. As an index of the shear strength, the internal friction angle reflects the friction property of rock and is an important engineering design parameter. From the results in Figure 16, according to the improved Mohr-Coulomb strength criterion, the internal friction angle value had a large discreteness, and the internal friction angle value was stable at about 25°.

Conclusions
In this paper, an experimental investigation was conducted on the mechanical parameters of CGFACPB cored from an underground coal mine. e main results are summarized as follows: (1) e ingredients of CGFACPB materials were analyzed, and the production process was introduced. e primary ingredients of CGFACPB materials include coal gangue, fly ash, OPC, and water. e field recipe of CGFACPB was described as follows: the ratio of OPC : fly ash : coal gangue was 1 : 4 : 6, and the solid concentration was 74%. Simultaneously the in situ coring layout plan and test procedures were provided. is study can provide significant reference for not only the evaluation of the supporting properties of CGFACPB in the field but also the design of the optimal CPB recipe. However, the field CGFACPB material is under the threedimensional stress, and there may be a larger difference in the results of mechanical properties from the unconfined compression test and the triaxial compression test. Currently, it is very difficult to measure the confining pressure of CGFACPB on the site. e next study may be focused on the investigation on the mechanical properties of CPB under triaxial loading test.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare no conflicts of interest.  Advances in Civil Engineering