Case Study on Pre-Splitting Blasting Reasonable Parameters of Goaf-Side Entry Retained by Roof Cutting for Hard Main Roof

Abstract

The key parameters of pressure relief pre-splitting blasting technology (PRPBT) contribute to the implementation effect of goaf-side entry by roof cutting (GSERC). This study is an attempt to design the key parameters including the roof cutting height, the roof cutting angle, the spacing between blasting boreholes, the charge structure and the sealing length by using theoretical analysis, numerical simulation and field implementation. The basic quality index of the main roof is 666.27, and combined with the peeping observation result, the main roof belongs to the category of a hard main roof. Different from a weakened roof and a compound roof, the unreasonable parameters of the PRPBT for the hard main roof lead to three problems including an insufficient pre-splitting blasting effect, roof suspension in the goaf and serious damage to the rock mass for the roof cutting rib. The PRPBT effect is closely related to key parameters, including the roof cutting parameters and the pre-splitting blasting parameters. Hence, to solve those three problems, a new design strategy was proposed based on the optimized directions of increasing the roof cutting height, decreasing the spacing between the blasting boreholes, changing the charge structure and adjusting the sealing length. According to the results of the theoretical calculation, numerical simulation and in situ measurement, the roof cutting height, the roof cutting angle, the spaced distance between the blasting boreholes, the charge structure and the sealing length were determined as 10 m, 15°, 500 mm, 4 + 4 + 3 + 3 + 2 and 2.4 m, respectively. The fissure rate was proposed to assess the developed effect of different pre-splitting blasting schemes. The proposed design strategy achieved a better effect of pre-splitting blasting for the entry, and eliminated the roof suspension of the hard main roof in the goaf. The design strategy in this study could contribute to other similar coal mines under the condition of a hard main roof.

1. Introduction

In response to the call of the national building energy-saving society, some innovative mining technologies have been applied aiming at high recovery rates, saving coal resources and low excavation [1]. Non-pillar coal mining (NPM) is one of the innovative technologies that have been widely used [2,3]. Regarding NPM, the goaf-side entry automatic retaining technology by roof cutting (GSERC) is a unique layout mode in which a coal pillar with a certain width is abolished between two panels [4]. The retaining technology for the GSERC includes pressure relief pre-splitting blasting technology (PRPBT), hydraulic fracturing technology, liquid carbon dioxide cracking technology, roof cutting technology with a chain arm saw and so on. Among these retaining technologies, pressure relief pre-splitting blasting is a dominant method because of its simple implementation craft, easy operating mode and lower cost [5]. The principle of PRPBT is the fracturing of rock mass in the set direction by using the energy accumulation effect. A series of new challenges arose in the applied process of the PRPBT, such as the large extrusion deformation of the gangue rib, the suspended roof with a large area in the goaf and the gangue retaining supporting structure being broken. In field investigation, these challenges have seriously threatened the safety of the workers and the equipment.

In recent years, fruitful results including field investigation, theoretical analysis and numerical simulations have been obtained. The optimal pre-splitting hole spaced distance based on the theory of bilateral cumulative tensile explosion was obtained [6]. The deep, shallow hole combined with cumulative energy blasting technology was proposed [7]. The mechanism of surrounding rock deformation in high-stress roadways was analyzed and a directional roof cutting technology was proposed [8]. The critical parameters design method on roof cutting pressure releasing goaf-side entry retaining technology including roof cutting design and blasting design was summarized [9]. A new technology of roof cutting with a chain arm was put forward [10]. The relationship between the roof cutting angle and the length of the cantilever beam was elaborated [11]. Moreover, the controlled blasting technique aiming at overbreaking of the underground mining development was also proposed [12]. Moreover, the pre-splitting blasting key parameters such as the roof cutting angle, the roof cutting height, the roof cutting hole spaced distance and others were studied [13].

It is clear that the effect of the PRPBT is not only closely related to the key parameters of pre-splitting blasting, but also deeply affected by the mining geological conditions such as the overlying strata features, the thickness of the main roof and the dip angle of the coal seam. Some findings have afforded us a reference in terms of a deeper understanding of the PRPBT of the GSERC under the condition of the composite roof. However, few studies have reported the PRPBT under the geological condition of the hard main roof. The hard main roof such as that of fine sandstone, coarse sandstone and other hard rock strata, was widely distributed in the Carboniferous period. The proportion of hard rock strata accounts for more than 60% [14,15,16]. Field investigation results indicated that the suspended roof occurred in the goaf after pre-splitting blasting [17]. The large load caused by the suspended roof can transmit to the surrounding rock mass and then the corresponding advanced abutment pressure and lateral abutment pressure make the stress environment more complex. Once the suspended roof loses stability, though dynamic ground pressure behaviors such as rock bursts, large deformation of surrounding rocks arises. Moreover, in the suspended roof area, the air easily flows into the goaf and then increases the probability of spontaneous ignition of the remaining coal mass. It is clear that under the geological condition of the hard main roof, the resulting dynamic ground responses caused by the hard main roof significantly deteriorate the supporting structures and even cause the appearance of roof collapse and large extrusion of the rib. Hence, it has become an urgent problem to be solved in the safe development process of coal resources.

To solve the problem of the suspended roof for the hard main roof, this study presented an in situ investigation of the pre-splitting blasting technology of the GSERC for the hard main roof, which is located in Dadougou coal mine, Datong city, Shanxi Province, China (Figure 1). The study is organized as follows: In Section 2, the original scheme and effect of the PRPBT are presented and the mechanical properties of the hard main roof are systematically investigated. In Section 3, the key parameters of the PRPBT are determined by using the theoretical analysis, numerical simulation and field test. Furthermore, a borehole peeping method is applied to examine the effect of different schemes, and then the optimized parameters are determined according to their effect. In Section 4, the effect of the optimized parameters on the suspended roof and the ground response is obtained. The results of the PRPBT key parameters and corresponding schemes can provide technical reference and theoretical guidance for other similar engineering practices.

2. Engineering Background

2.1. Geological Conditions

Dadougou coal mine adopted the NPM technology, the mineable coal seam is 2# with an average thickness of 2.4 m and a dip angle of 3°. The burial depth of coal seam 2# is approximately 480 m. The geological survey report indicated the immediate roof is silty mudstone with an average thickness of 2.0 m, the main roof is coarse sandstone with an average thickness of 13.6 m and the floor is mudstone and fine sandstone with an average thickness of 0.87 m and 5.8 m, respectively. The geological stratigraphy column of coal seam 2# is presented in Figure 2.

This study was conducted in tailgate 5201, with a length of 1726 m. The tailgate 5201 was used for transportation and was pedestrian in its primary used stage; after panel 5201 was mined out, the tailgate 5201 was retained as the headgate 5202 of panel 5202 for ventilation (Figure 3). The retained mode of tailgate 5201 is along the goaf-side by adopting the PRPBT. The height and width of the tailgate 5201 are 3.6 m and 5.2 m, respectively.

2.2. Rock Mechanical Properties of Hard Main Roof

To better determine rock mechanical properties of the roof for coal seam 2#, laboratory tests including the modulus of elasticity, the uniaxial compressive strength, the tensile strength and the cohesion were conducted. The rock samples were collected from silty mudstone and coarse sandstone. The experiments were implemented with a servo-controlled testing system (MTS815.03). A total of 60 rock cores were tested by conducting the uniaxial tensile experiment and the uniaxial compressive experiment. To obtain better test results, the rock cores were divided into three groups. According to the test results, the average values of mechanical parameters are illustrated in

Table 1. Rock mechanical property experimental results.

Strata	Lithology	Ei (GPa)	σc (MPa)	σt (MPa)	C/MPa
Roof	Coarse sandstone	78.96	117.09	3.86	24.25
	Silty mudstone	39.87	35.79	2.91	16.38

E_i is the modulus of elasticity; σ_c is the uniaxial compressive strength; σ_t is the tensile strength; C is the cohesion.

.

In combination with the rock mass classification standard [18,19,20], the basic quality index of rock mass was calculated for rock mass classification.

$$BQ=90+3R_C+250K_v$$

(1)

where BQ is the basic quality index of rock mass; $R_C$ is the uniaxial compressive strength of rock mass; $K_v$ is the rock mass integrity index value. $K_v$ can be determined according to

Table 2. Comparison between J_V and K_V.

Volume Joint Number of Rock Mass Jv (Number/m3)	<3	3~10	10~20	20~35	>35
Kv	>0.75	0.75~0.55	0.55~0.35	0.35~0.15	<0.15

.

The field investigation of the joint number of the silty mudstone and the coarse sandstone was conducted in different regions. According to the field investigation results, the average values of the volume of the joint number of rock mass for the silty mudstone and the coarse sandstone were 8.7/m³ and 2.3/m³, respectively. Hence, the corresponding rock mass integrity index value could be determined as 0.65 and 0.90, respectively. Combined with the results of uniaxial compressive strength, the basic quality indexes of the silty mudstone and the coarse sandstone were 422.37 and 666.27, respectively. Following China’s engineering rock mass classification standard (GB50218-2014), the basic quality grades of the silty mudstone and the coarse sandstone were class III and class I, respectively. Consequently, the immediate roof belongs to the category of the relatively hard roof, while the main roof belongs to the category of the hard roof.

To verify the accuracy of the calculation results, three peeping boreholes with a diameter of 42 mm and a length of 15 m were drilled to observe the integrity of the main roof. The fracture development in the roof was detected by borehole peeping equipment with a type of CKK12A, which consists of a host machine, a focusing camera, a peeping depth recorder and a connecting rod. It can record a view of fracture development along the depth of borehole in real time. The view can be transmitted to the host machine during observation, and then the fracture development can be analyzed by imaging software.

The boreholes were arranged before the retreating of panel 5201. Borehole 1# was arranged in the middle span of the roof and boreholes 2# and 3# were arranged 1.5 m away from the rib, respectively. The peeping results and their corresponding analysis are illustrated in Figure 4. Regarding borehole 1#, the vertical fracture developed at the depth of 2.2 m and the rock mass was intact at the depth of greater than 3.5 m. As for borehole 2#, the bedding separation developed at the depth of 2.5 m and the vertical fracture emerged at the depth of 3.7 m. However, the intact rock mass developed at the depth of greater than 3.7 m. As shown in Figure 4c, the vertical fractures developed at depths of 1.7 m and 3.3 m, respectively. The rock mass was integral at depths greater than 3.3 m. It is concluded that the vertical fracture and the bedding separation mainly developed in the shallow roof region within 3.7 m. When the depth was greater than 3.7 m, the rock mass was intact. Visually, the excavation of tailgate 5201 caused the breaking of the shallow roof region but had little impact on the deep roof. Because of the hard main roof, the rock mass was intact.

3. Initial Parameters and Implementation Effect of the PRPBT

3.1. Initial Parameters of the PRPBT

The initial parameters of the PRPBT can be summarized as follows: (1) The energy-gathered tube with two specified directions was adopted in the process of roof cutting. The external diameter, internal diameter and the length of the tube were 42 mm, 36.5 mm and 1500 mm, respectively. (2) The tertiary emulsion explosive had a diameter of 35 mm and a length of 300 mm. Several tertiary emulsion explosives were embedded in the energy-gathered tube according to the charge structure. The initial charge structure was 3 + 3 + 2 + 2 + 1. The blasting boreholes were sealed by the yellow mud and the sealing length was 2.5 m. (3) The roof cutting length was 8.1 m and the roof cutting angle was 15°. The spaced distance between the blasting holes was 800 mm.

3.2. Implementation Effect of the PRPBT

3.2.1. Insufficient Pre-Splitting Blasting Effect

The roof cutting effect is closely related to the roof cutting parameters, for example, the roof cutting height, the roof cutting angle, the sealing length for blasting boreholes, the charge structure, the ignition mode and the ignition number of boreholes with each blasting [21]. A field investigation including in situ measurement and borehole peeping was conducted to explore the roof cutting effect. The boreholes’ peeping results indicated that the pre-splitting blasting effect was poor, as directional fissures were not developed and the rock mass in some regions was not cracked, as shown in Figure 5.

3.2.2. Roof Suspension in the Goaf of Panel 5201

The roof cutting length and the charge structure are two important factors affecting the effect of the roof cutting and the roof collapse. The main roof of coal seam 2# belongs to the kind of hard roof, and the roof structure is the structure of “immediate roof + hard main roof”. Regarding the hard main roof, the roof cutting length and the charge structure should be determined based on mastering its feature. Under the condition of adopting the roof cutting height of 8.1 m and the charge structure of 3 + 3 + 2 + 2 + 1, the pre-splitting blasting effect is poor. The charge structure of 3 + 3 + 2 + 2 + 1 means that five energy gathered tubes were filled in order along the borehole depth. Regarding the charge structure, three explosives were embedded in the first energy-gathered tube, three explosives were embedded in the second energy-gathered tube, two explosives were embedded in the third energy-gathered tube, two explosives were embedded in the fourth energy-gathered tube and one explosive was embedded in the fifth energy-gathered tube. The explosive adopted three-stage emulsion explosives permitted in coal mines with a diameter of 35 mm, a length of 30 cm and a mass of 300 g. The energy-gathered tubes were connected by a millisecond delay electric detonator.

Several roof suspensions occurred in the goaf and the surrounding rock mass of their corresponding entry region was seriously deformed. The location and suspended area of the roof suspension was field measured by a laser range finder along the axial direction of tailgate 5201. The distance from the current open-off cut, the suspended length and the distance from the roof cutting side were all obtained (

Table 3. Statistics of the roof suspension in the goaf of panel 5201.

Order Number	Distance from the Current Open-Off Cut (m)	Suspended Length (m)	Distance from the Roof Cutting Side (m)	Suspended Area (m2)
(a)	129	18	8	144
(b)	251	14	2~4	28~64
(c)	290	22	2.5	55
(d)	300	25	10	250
(e)	330	12	6	72
(f)	415	10	3	30
(g)	540	13	5~6	65~78
(h)	667	10	4~5	40~50

). It can be seen that the maximum length of the roof suspension is 27.3 m and the maximum depth from the roof cutting side is 10.8 m, and the maximum area of the roof suspension can reach a value of 250 m². There are nine roof suspensions in the goaf with a range from 0 m to 700 m; the details are illustrated in Figure 6.

3.2.3. Serious Damage of the Rock Mass for Roof Cutting Rib

The sealing length for pre-splitting blasting boreholes is an important factor affecting the blasting effect. The actual scene of the roof cutting rib damage is shown in Figure 7. When adopting the sealing length with a value of 2.5 m, the roof near the roof cutting side developed fissures and thus caused the potential roof collapse (Figure 7a). Moreover, the metal mesh that was used to protect the roof cutting rib was seriously destroyed. It wasted the supporting materials and added to the supporting expense. Additionally, the rock mass of the roof cutting rib was seriously damaged, as shown in Figure 7b. The anchor bolts were destroyed and then the anchoring action was lost.

4. Optimized Direction of the PRPBT

According to the field investigation results, it can be concluded that the unreasonable parameters of the PRPBT had a poor implementation effect. The results from existing research indicate that the design parameters of pre-splitting blasting and roof cutting have a significant impact on the roof cutting effect [22,23]. The pre-splitting blasting parameters include the per blasting borehole charge quantity, the blasting boreholes’ spaced distance and the single detonating quantity. Regarding the roof cutting parameters, these consist of the roof cutting height and roof cutting angle. Hence, to achieve a better effect, these parameters should be determined according to the characteristics of the hard main roof. Combined with previous results of the PRPBT and the successful support experiences in many underground coal mines in China, the following optimized directions are proposed for tailgate 5201:

➢

Increase the roof cutting height. We acquired a hard main roof with a thickness of 13.6 m. Under the condition of the roof cutting length with a value of 8.1 m and the roof cutting angle with a value of 15°, the roof cutting height was 7.82 m. It means that the roof cutting action mainly occurred in the lower part of the hard main roof, so when the lower part collapsed insufficiently, the upper part of the hard main roof was easily suspended. Technically, the greater the roof cutting height, the better the effect of the roof cutting. The roof cutting height should cover the upper part of the hard main roof as soon as possible.

➢

Decrease the spaced distance between the blasting boreholes. The spaced distance between the blasting boreholes is an important factor affecting the blasting effect [9]. Generally speaking, the smaller the spaced distance between the blasting boreholes, the easier the crack penetrates; hence, the better the pre-splitting blasting effect. When the spaced distance between blasting boreholes is 800 mm, roof suspensions occurred in the goaf. Hence, to eliminate the roof suspension, the spaced distance between the blasting boreholes should be decreased.

➢

Change the charge structure. It has been indicated that the charge structure has an important influence on the explosive energy, explosive pressure and rock mass damage [9,24]. If the charge structure is unreasonable, the pre-splitting blasting effect will be poor. Accordingly, designing a reasonable charge structure can improve the effect of pre-splitting blasting.

➢

Adjust the sealing length. The sealing length can be divided into two conditions according to the pre-splitting blasting effect. If the sealing length is too long, the utilization efficiency of boreholes decreases. If the sealing length is too short, the explosive gas rushes out of the borehole, and the shock wave forms [25]. The energies induced by the shock wave can destroy the rock mass near the roof cutting side and other supporting structures such as the metal mesh (Figure 7). It is necessary to determine the reasonable sealing length.

5. New Strategy of the PRPBT and Its Field Application

5.1. New Design Parameters of Roof Cutting

5.1.1. Roof Cutting Height

Under the condition of the hard main roof, the design of the roof cutting height should consider two aspects: first, after the rock mass within the roof cutting range collapses, it should be able to fill the goaf at the roof cutting side of GSERC and form effective support for the roof in the goaf; second, if the pre-splitting blasting effect is poor, the roof suspension occurs easily in the goaf. Hence, the roof cutting height should be reasonable for preventing roof suspension. The roof cutting height is mainly calculated by the swelling feature of the gangue. Combined with the existing research results [26], the calculation formula of the cutting height is given as follows:

$$H_{QF}=(H_M-\Delta H_1-\Delta H_2)/(K-1)$$

(2)

where $H_{QF}$ is the roof cutting height, m; $H_M$ is the mining height, m; $\Delta H_1$ is the roof subsidence, m; $\Delta H_2$ is the floor heave, m; and $K$ is the swelling coefficient.

According to the measured parameters, the mining height of coal seam 2# is 3.0 m. The average value of roof subsidence and the floor heave is 0.2 m and 0.1 m, respectively. Based on the empirical value, the swelling coefficient of coarse sandstone is determined as 1.28. The calculation result of the roof cutting height is 9.64 m. To verify the rationality of the theoretical calculation, a numerical simulation was conducted to explore the vertical displacement changeable law of the rock mass under the roof cutting heights of 8 m, 9 m, 10 m and 11 m by using the UDEC 4.0 software, as shown in Figure 8. UDEC discrete element numerical simulation software fully considers the stress and strain situation of rock mass. Furthermore, after analyzing the contact relationship between rock blocks, the interaction between the structural plane and rock block can be truly reproduced and the mutual dislocation, collapse instability or deformation failure of the rock block can be vividly reproduced.

When the roof cutting height was 8 m, the immediate roof collapsed directly, but the immediate roof was not fully filled with the space between the main roof and the floor because the main roof was not completely cut off. The key blocks were hinged with each other, which then resulted in the state of a cantilever beam. Under the loading action of overlying strata and itself, the rotary movement of the main roof occurred, and then had a serious influence on the surrounding mass of the GSERC. Hence, vertical displacement ranging from 1.5 m to 2.0 m developed in the roof of the GSERC, as shown in Figure 8a. Compared with the roof cutting height of 8 m, vertical displacement ranging from 1.0 m to 2.0 m developed in the roof near the roof cutting side. When the roof cutting height was 9 m, the immediate roof also collapsed directly, but due to insufficient roof cutting action, the main roof had an adverse influence on GSERC. It can be obtained that the roof near the roof cutting rib had a serious subsidence. This situation is not conducive to GSERC maintenance.

When the roof cutting height was 10 m, the immediate roof also collapsed directly and it was able to fully fill the space between the main roof and the floor. In this condition, the immediate roof could provide a supporting act for the main roof, and then the main roof was in a stable state. Correspondingly, the main roof movement had little effect on the GSERC because it had been cut off, and this brought a beneficial environment for the GSERC support. A vertical displacement with a value less than 0.5 m developed in the roof near the roof cutting side. As for the roof cutting height of 11 m, it also has a similar feature. Therefore, considering the economic cost and technology implementation, the roof cutting height was determined as 10 m.

5.1.2. Roof Cutting Angle

The design of the roof cutting angle mainly considers the two following aspects. First, a reasonable roof cutting angle can minimize the lateral vibration on the roof near the roof cutting side of the GSERC when the roof collapses. Second, the roof cutting angle should try to keep the GSERC stress environment in a low state so as to facilitate its maintenance. Based on the existing research results [27], the calculation formula of the roof cutting angle is given as follows:

$$\beta =\frac{\pi }{2}-\arctan \frac{L_Y-L_H}{H_M}$$

(3)

where $\beta$ is the roof cutting angle, (°); $L_Y$ is the periodic weighting distance, m; $L_H$ is the roadway width, m; and $H_M$ is the mining height of coal seam 2#, m.

According to the measured data, the periodic weighting distance of 8201 working face is 18 m, the roadway width is 5.2 m and the maximum mining height of coal seam 2# is 3 m. The calculation result of the roof cutting angle was 13.2°. Based on the operation conditions and the relevant operation experience, the roof cutting angle was determined as 15°.

5.1.3. The Spaced Distance between the Blasting Boreholes

LS-DYNA 17.0 software is a fully functional program for geometric nonlinearity, material nonlinearity and contact nonlinearity. In the process of pre-splitting blasting, the damage caused by the detonation wave to the surrounding rock belongs to the typical nonlinear damage, which is characterized by complex anisotropic damage. LS-DYNA software can effectively simulate the crack propagation of a rock mass under the action of high-energy explosives. To determine the spaced distance between the blasting boreholes, the numerical simulation model including two boreholes was established by using LS-DYNA software. Numerical simulation models for the spaced distance between the blasting boreholes with a value of 400 mm, 500 mm and 600 mm were established, respectively. The model was established by the SOLID64 element, and the fluid–solid coupling algorithm was used to simulate the damage to the rock mass caused by blasting. In the process of numerical simulation, the ALE grid was used to simulate explosive charges and air, while the Lagrange grid was used to simulate the rock mass and the energy gathering tube. The rock mass, energy gathering tube, explosive charges and air were coupled and the fluid–solid coupling model was established by the simulation language with * CONSTRAINED-LAGRANGE-IN-SOLID. The simulation language with * MAT_HIGH_EXPLOSIVE_BURN was used to define the material model of the tertiary emulsion explosive. The detailed parameters of the mining emulsion explosive, air and fine sandstone are presented in

Table 4. Material parameters of mining emulsion explosive.

Density (kg·m−3)	Detonation Velocity (m·s−1)	Detonation Pressure (GPa)	A (GPa)	B (GPa)	r1	r2	$\mathit{\xi }$	e (GPa)
1175	3100	3.0	252	10.28	7.22	2.38	0.064	2.58

,

Table 5. Air material parameters.

Density (kg·m−3)	C0	C1	C2	C3	C4	C5	C6	V0	E0
1.293	−1 × 10−6	0	0	0	0.38	0.38	0	1.0	2.059 × 10−6

and

Table 6. Physical and mechanical parameters of fine sandstone.

Lithology	Density (g·cm−3)	Tensile Strength (MPa)	Uniaxial Compressive Strength (MPa)	Elastic Modulus (GPa)	Poisson’s Ratio	Cohesion (MPa)	Friction Angle (°)
Fine sandstone	2.52	3.57	46.7	68.8	0.165	26.8	41

. Explosive pressure was expressed by the equation of state [9]:

$$Q=\mathrm{a}\left(1-\frac{\xi }{a{r}_1}\right)e^{-r_{1}v}+b\left(1-\frac{\xi }{a{r}_2}\right)e^{-r_{2}v}+\frac{\xi {\rho }_0}{v}$$

(4)

where Q is the explosive pressure, MPa; a, b, $r_1$, $r_2$, $\xi$ are the material parameters of the emulsion explosives; v is the volume of the explosive products during the explosion; and ${\rho }_0$ is the density of the explosive products.

The crack development process under the spaced distance between two blasting boreholes with values of 400 mm, 500 mm and 600 mm was presented. The crack development process was divided into the crack formation stage, crack extension stage, crack coalescence stage and crack expansion stage. When the spaced distance between two blasting boreholes was 400 mm, with the propagation of explosive-induced gas, the crack continues to extend in the rock mass along the pre-splitting direction and the cracks of the two holes intersect each other to produce new cracks, which connect two blasting boreholes. The rock mass between two blasting boreholes can be penetrated; moreover, the crack can be expanded (Figure 9).

When the spaced distance between two blasting boreholes was 500 mm, the rock mass between two blasting boreholes just penetrated (Figure 10). Compared with the case of 400 mm, the directional crack propagation needed a longer process when the spaced distance was 500 mm. When the spaced distance is 500 mm, the crack extension can be divided into three stages: crack initiation stage, continuous extension stage and extension sudden increase stage. At the crack initiation stage, the crack extends slowly and only extends a certain distance around the borehole wall. As the explosive-induced gas continues to break the rock mass, the crack continues to extend and enters the continuous extension stage. Subsequently, the crack enters the stage of sudden increase in extension and quickly penetrates two blasting boreholes. When the spaced distance between two blasting boreholes is 600 mm (Figure 11), the rock mass cannot be penetrated. According tothe above analysis, considering economic cost, the spaced distance between two blasting boreholes was determined as 500 mm.

5.2. New Design Parameters of Pre-Splitting Blasting

In the engineering practice of NPM, the core technology of the PRPBT is bidirectional energy concentrated pre-splitting blasting technology [6]. As a kind of pre-splitting blasting technology, this technology can realize the directional transmission of explosive energy along the direction of the shaped tube through the combination of the shaped charge tube and the explosive and then realize the directional fracture of rock mass along the roof cutting direction.

The determined parameters of pre-splitting blasting including charge structure, sealing length and the number of single initiations were mainly determined by field investigation. Existing results have presented that the pre-splitting blasting effect is judged by indirect indexes such as gangue retaining pressure, the periodic weighting length and the loading of a single hydraulic prop near the gangue rib [28]. The measured result of the indirect index could grasp the roof cutting effect, but it cannot form a systemic design method for the key parameters of pre-splitting blasting. Therefore, the fissure rate along the borehole length is proposed to design the parameters of pre-splitting blasting and assess its effectiveness. The obtained process of the fissure rate can be described as follows: first, the video of roof fracture development along the borehole depth was obtained. Second, the number of the developed fracture zones including lateral fissures, vertical fissures, bedding separations and annular fractures was recorded per meter. Based on the result of the number of developed fracture zones, the lengths of the developed fracture zones were analyzed. Third, the length of the developed fracture zone per meter is called the fissure rate. Hence, the fissure rate is calculated by the observation result, it can be expressed by the following equation:

$$L_{\mathrm{n}}=\frac{F_n}{1}\times 100\%$$

(5)

where $L_{\mathrm{n}}$ is the crack rate within one meter along the borehole length, %; and $F_n$ is the crack length within one meter, m.

Based on the successful practice of the PRPBT in many coal mines, this study adopted a dynamic adjustment strategy regarding the single borehole. The strategy was described as follows: (1) A single borehole charge structure (Figure 12) was determined according to the borehole length. In the meantime, the sealing length was also obtained. (2) Then field tests were conducted according to the design strategy. The borehole peeping equipment was used to view the crack development along the borehole direction. The fissure rate was calculated based on the peeping result of crack development. (3) Based on the fissure rate, the charge structure and the sealing length were adjusted dynamically. The single borehole pre-splitting blasting test was conducted again according to the charge structure until the demand for the fissure rate was met for the PRPBT. The pre-splitting blasting strategies are presented in

Table 7. The pre-splitting blasting strategy.

The Strategy	The Spaced Distance between Two Blasting Boreholes (mm)	The Filled Structure of Energy Gathered Tube	Charge Structure	The Sealing Length (m)
1	500	1.5 + 1.5 + 1.5 + 1.5 + 0.5	3 + 3 + 3 + 2 + 1	2.7
2	500	1.5 + 1.5 + 1.5 + 1.5 + 0.5	3 + 3 + 3 + 3 + 1	2.7
3	500	1.5 + 1.5 + 1.5 + 1.5 + 0.5	4 + 3 + 3 + 2 + 1	2.7
4	500	1.5 + 1.5 + 1.5 + 1.5 + 1.5	4 + 3 + 3 + 3 + 1	2.4
5	500	1.5 + 1.5 + 1.5 + 1.5 + 1.5	4 + 3 + 3 + 3 + 2	2.4
6	500	1.5 + 1.5 + 1.5 + 1.5 + 1.5	4 + 4 + 3 + 3 + 2	2.4

.

In the field test process, an energy-gathering tube with a length of 1.5 m, an outer diameter of 42 mm and an inner diameter of 36.5 mm was adopted. A three-grade emulsion explosive roll with a diameter of 35 mm and a length of 300 mm was used. Explosive rolls were connected by a millisecond delay detonator. After the pre-splitting blasting works of each strategy were finished, the borehole peeping was conducted to observe the fissure development.

The peeping results of each strategy are shown in

Table 8. The peeping results of each strategy.

Order Number	The Strategy	Peeping Results of the Fissure Development for Blasting Boreholes
		3.0 m	5.0 m	7.0 m	9.0 m
A	1
B	2
C	3
D	4
E	5
F	6

. The analysis of the peeping results along the borehole lengths of 3.0 m, 5.0 m, 7.0 m and 9.0 m was obtained. Regarding the charge structures of 3 + 3 + 3 + 2 + 1 and 3 + 3 + 3 + 3 + 1, in the shallow section within the borehole length of 3.0 m, no fissures developed (Table 8A,B). This was because the sealing length was 2.7 m and the explosion-induced blasting energies had difficulty acting effectively in this section. At the borehole length of 5.0 m, fissures occurred in the borehole, but there were no bidirectional fissures. At the borehole lengths of 7.0 m and 9.0 m, bidirectional fissures occurred in the borehole. It can be concluded that the bidirectional fissures mainly developed in the deep section of the borehole and the pre-splitting blasting effect was relatively poor in the shallow section. Based on the charge structures of 3 + 3 + 3 + 2 + 1 and 3 + 3 + 3 + 3 + 1, the explosive rolls were added in the deep section, then the charge structure of 4 + 3 + 3 + 2 + 1 was formed. As shown in Table 8C, no fissures occurred in the sealing section of the borehole; the bidirectional fissures occurred in the sections along the borehole lengths of 5.0 m and 7.0 m. At the borehole length of 9.0 m, fissures occurred in three directions and the rock mass was cracked. Through the analysis of the peeping results for the charge structures of 3 + 3 + 3 + 2 + 1, 3 + 3 + 3 + 3 + 1 and 4 + 3 + 3 + 2 + 1, it can be seen that the pre-splitting blasting is poor in the shallow section of the borehole.

Compared with the charge structures of 3 + 3 + 3 + 2 + 1, 3 + 3 + 3 + 3 + 1 and 4 + 3 + 3 + 2 + 1, to obtain a better pre-splitting blasting effect at the shallow part, the sealing length was decreased to 2.4 m and the explosive rolls were added. The charge structures of 4 + 3 + 3 + 3 + 1, 4 + 3 + 3 + 3 + 2 and 4 + 4 + 3 + 3 + 2 were proposed. As for the charge structure of 4 + 3 + 3 + 3 + 1, the bidirectional fissures developed at the borehole length of 3.0 m and 5.0 m, while the fissures with three directions occurred at the borehole lengths of 7.0 m and 9.0 m. Concerning the charge structures of 4 + 3 + 3 + 3 + 2 and 4 + 4 + 3 + 3 + 2, it can be seen that the bidirectional fissures developed in the whole section of the borehole and a better pre-splitting blasting effect was achieved. As for the charge structure of 4 + 4 + 3 + 3 + 2, the dimensions of the bidirectional fissures were greater than the charge structure of 4 + 3 + 3 + 3 + 2, and the pre-splitting blasting effect was better. It is concluded that the charge structure of 4 + 4 + 3 + 3 + 2 can achieve a better pre-splitting blasting effect than the other charge structures.

To further assess the pre-splitting blasting effect of different charge structures, the fissure rate was also obtained. As shown in Figure 13, the maximum value of the fissure rate for the charge structures of 3 + 3 + 3 + 2 + 1, 3 + 3 + 3 + 3 + 1 and 4 + 3 + 3 + 2 + 1 were 20%, 23% and 21% when the borehole length was less than 3.0 m, respectively. However, the maximum value of the charge structures of 4 + 3 + 3 + 3 + 1, 4 + 3 + 3 + 3 + 2 and 4 + 4 + 3 + 3 + 2 increased to 86%, 90% and 92%, respectively. The fissure rates of the charge structures of 4 + 3 + 3 + 3 + 1, 4 + 3 + 3 + 3 + 2 and 4 + 4 + 3 + 3 + 2 were greater than the charge structures of 3 + 3 + 3 + 2 + 1, 3 + 3 + 3 + 3 + 1 and 4 + 3 + 3 + 2 + 1. The results of the fissure rate were consistent with the peeping results. Hence, the charge structure was determined as 4 + 4 + 3 + 3 + 2.

5.3. Field Application

According to the results of the new design parameters of the PRPBT, it can be acquired that the roof cutting height was 10.0 m, the roof cutting angle was 15°, the charge structure was 4 + 4 + 3 + 3 + 2 and the sealing length was 2.4 m. Additionally, to meet the requirement of panel 5201 coal production, the daily mining progress should be greater than 4.8 m, namely, the daily pre-splitting blasting progress was 4.8 m/d. Then there were eleven boreholes that needed to be blasted every day. Considering the work arrangement, three boreholes were blasted each time and four blast times were arranged.

To evaluate the performance of the newly designed pre-splitting blasting strategy, comprehensive field observations, including borehole peeping observation and the field measurement of roof suspension, were conducted in the tailgate 5201. The borehole peeping observation was measured using borehole peeping equipment with the mode of CXK12(A). The development of roof suspension was measured by flexible tape and a laser range finder. After the newly designed pre-splitting blasting strategy was implemented, the field observations were conducted within 200 m behind the goaf. The observation results regarding the fissures’ development are presented in Figure 14. As shown in Figure 15, the bidirectional fissures developed in the blasting borehole.

Based on investigating the peeping results of pre-splitting blasting, the roof suspension in the goaf was also investigated on site and the measured range was within 200 m behind the goaf. There are only three roof suspensions, and the maximum value of the roof suspension was only 3.2 m², which was about 0.1 times the average value before using the newly designed strategy. After adopting the newly designed strategy, the hard main roof collapsed fully in the goaf, as shown in Figure 15. It can be seen that the newly designed strategy achieves a better pre-splitting blasting effect and can also significantly reduce the roof suspension in the goaf. Notably, this study is based on the specific condition of Dadougou coal mine; the results can provide technical reference and theoretical guidance for other coal mines with similar conditions.

6. Conclusions

This study implemented in Dadougou coal mine, Datong city, Shanxi province, China, aimed to study the pre-splitting blasting technology of PRPBT for a hard main roof, based on field investigations, numerical simulations and theoretical analysis. The conclusions contain the following three aspects:

(1)

The laboratory tests indicated that the elasticity modulus, the uniaxial compressive strength, the tensile strength and the cohesion of the coarse sandstone were 78.96 GPa, 117.09 MPa, 3.86 MPa and 24.25 MPa, respectively, and the basic quality was 666.27. The peeping observation results presented that the vertical fissure and the bedding separation mainly developed in the shallow roof section within 3.7 m, but the rock mass was intact when the depth was greater than 3.7 m. According to the comprehensive result, coarse sandstone belongs to the category of hard roof.

(2)

When adopting the original parameters of the PRPBT, an insufficient pre-splitting blasting effect, roof suspension in the goaf and serious damage of the rock mass for the roof cutting rib emerged. Increasing the roof cutting height, decreasing the spaced distance between the blasting boreholes, changing the charge structure and adjusting the sealing length were the optimized directions of the PRPBT for the hard main roof.

(3)

A newly designed strategy of the PRPBT incorporating the design of roof cutting parameters and pre-splitting blasting parameters was proposed. Based on the comprehensive results of the theoretical analysis, numerical simulation and field implementation, the new parameters including the roof cutting height, the roof cutting angle, the spaced distance between the blasting boreholes, the charge structure and the sealing length were determined as 10 m, 15°, 500 mm, 4 + 4 + 3 + 3 + 2 and 2.4 m, respectively. Field practice results indicated that the newly designed strategy could crack the rock mass in two directions and achieved a better pre-splitting blasting effect. The maximum value of roof suspension was about 0.1 times the average value with the initial scheme and roof suspension reduced significantly. The PRPBT obtained a better pre-splitting blasting effect for the entry. The strategy can provide a guidance for similar mining conditions.