Optimization of Support Parameters for Large-Section Open-Cut Roadways in Fully Mechanized Mining with Large Mining Height

Abstract

The extensive adoption of large mining height technology and the progressive deepening of mining operations have presented formidable challenges to the safety of roadway support. The selection of roadway support configurations and their operational parameters is critically important in underground mining operations. Taking the open cut of Hongliu Coal Mine as the engineering background, this study conducts similar model experiments and field monitoring to evaluate the large-section open-cut support system. We aim to address unreasonable parameters and the low efficiency of this system in fully mechanized mining faces with large mining heights. The results demonstrate that deformation and failure initially occur at the cut corners. According to field observation data, the convergence of the system’s two sides across the three measuring stations is markedly greater than the roof subsidence on average (104.9 mm vs. 46.0 mm). This indicates the collapse of surrounding rocks on both sides. The peak abutment pressure of the cutting hole occurs approximately 16 cm from the coal wall (scaled to 3.2 m on site).

1. Introduction

Coal, as a fundamental energy source and industrial raw material, plays a dominant role in energy production and consumption. As China’s economy continues to develop, the demand for coal resources is increasing. With the depletion of shallow coal reserves, mining operations are inevitably transiting to deeper deposits. However, this transition brings about significant changes in the mining environment, with increased mining depth. Due to the complex impacts of “three highs and one disturbance”, the green, safe, and efficient production of deep coal is facing serious challenges [1,2,3,4,5]. One of the critical aspects affected by these changes is the support design for large-section open-cut roadways in fully mechanized mining faces with large mining heights. The opening cut provides the necessary place for the installation of mining equipment and serves as the site for working face mining. The stability of opening cut support plays a crucial role in the smooth installation of mining equipment and subsequent mining operations. In recent years, with the improvement in coal mining technology, mining has been continuously developed towards larger-scale, intensive, heavy-duty, and intelligent operations, especially the wide application of large mining height technology. In addition, as mine depths increase, the dimension of the opening cut in fully mechanized mining working face is continually increased. This expansion in the height and width of the opening cut roadway has resulted in larger deformation and fracture zones in the surrounding rock, making support more challenging and leading to increased risks of roof caving, roof falls, and dynamic impacts [6,7,8,9,10].

In 1956, China’s coal mines began to use anchoring rod support technology in rock roadway, with a history spanning over 60 years [11]. This technology has evolved from low strength to high strength, and further to high pre-stress and strong support [12,13,14,15,16]. In conditions of increased stresses, both vertical and horizontal, one of the best solutions for the support of large-size workings is cable anchors, which, thanks to their elasticity and flexibility, perfectly adapt to the deformation of rocks around the workings [17], and additionally demonstrate high dynamic load-bearing capacity [18]. Currently, anchoring rod support technology for coal mine roadway is being actively promoted and applied domestically, yielding favorable technical and economic benefits [19,20,21]. The large-section cut of a large mining height working face differs from the general thin and medium-thickness coal seam roadway, and features characteristics such as large section size, severe deformation of surrounding rock, rapid early-stage deformation, large total volume, long durations, and overall rock mass deformation [22,23,24,25]. Inappropriate selection of support forms and parameters can lead to two extremes: excessive support strength resulting in material waste, or inadequate support strength resulting in roof falls [26,27,28,29]. Therefore, the optimal selection of anchoring rod support parameters for the large-section opening cut holds paramount importance, and many experts and scholars have conducted extensive research in this aspect [30,31,32,33,34].

Mining causes disturbance of the coal rock body, resulting in the generation of joints, fissures, and other discontinuous surfaces, thereby changing the strength characteristics of the coal rock body [35,36,37,38]. According to the three key factors determining roadway stability—the surrounding rock strength, surrounding rock stress, and support technology—research on the optimal anchoring rod support parameters for large-section open-cut hole has been carried out around these three aspects [39,40,41,42]. This study is of significant importance to achieve stability in supporting surrounding rock in deep and large-section opening cuts. This paper takes the 1121 working face of Hongliu coal mine as the research object, conducts a physically similar model experiment, and carries out research on the design of the opening cut support parameters of Hongliu coal mine around the stability of the surrounding rock of the large-section opening cut under a large mining height. As it is the initial mining face, the surrounding rock destructive characteristics of the working face have not been completely grasped. To ensure the safe, efficient, and early production of the initial mining face and to guide better support for the next working face, it is necessary to start from the combined anchoring rod and anchor cable support technology. Suboptimal parameters and operational inefficiencies persist in the support systems of large-section open off-cuts within high-cutting-height longwall faces. In this study, through numerical simulations and theoretical analyses, the critical parameters influencing support effectiveness in large-section open off-cuts were systematically identified. The feasibility of the proposed optimized support parameter scheme was verified by comparative analysis between field monitoring data and simulation results. This investigation provides scientific guidance for support design in large-mining-height working faces. This study advances prior research by conducting multi-parameter optimization through integrated analysis of similar material simulation experiments and field monitoring data, developing a comprehensive methodology for support parameter refinement. This approach incorporates systematic considerations of material selection and parametric configuration to evaluate the impact of support structure layouts on operational effectiveness, achieving synergistic optimization of critical parameters. Validation was ultimately conducted through practical engineering applications. This involved verifying, designing, and optimizing existing support schemes and proposing a set of both safe and cost-effective optimized support parameters based on on-site mining pressure observations. The results of the research can also serve as a reference for the comprehensive control of the support of large-section cuts in similar large-height working faces. This work significantly enhances the stability and safety of support systems through improved parameter optimization.

2. Engineering Overview

2.1. Geological Overview of the Working Face Cut

The Hongliu Coal Mine is one of the large mines planned and constructed within the Yuanyanghu Coal Mine of the Ningdong Coalfield, with a designed production capacity of 8.0 Mt/a. The 1121 working face is located on the eastern side of mining district 11, with a working face length of 300 m and an advance length of 1480 m. The thickness of the No. 2 coal seam ranges from 4.99 to 6.05 m, with an average thickness of 5.4 m, classifying it as a thick coal seam. As the initial mining face, large-mining-height mining technology is employed, with the mining height varying according to the coal seam thickness. The controlled caving of overlying strata above mined-out areas is achieved through natural subsidence processes or artificially induced techniques. To achieve modernized mining production, the working face utilizes 6.2 m high double-pillar shield-type hydraulic support from Zhengzhou Coal Machinery Factory. To facilitate the installation of a fully mechanized mining face, the mine starts from the essence of combined support using anchoring rods and anchor cables. This approach considers the characteristics of large-span coal roadways, and proposes a pre-stressed anchoring support technology plan for large-section coal roadway. The dimension of the cut is designed to be 8.6 × 4.2 m, classifying it as a large-section cut, and presenting significant support challenges. The traditional construction methods are no longer effective in controlling the deformation and damage of the surrounding rock.

2.2. Existing Support Forms for the Initial Cut

The opening cut is excavated along the roof of the coal seam, employing a support scheme that consists of anchor mesh and beam support combined with anchor cables. The designed net length of the opening cut is 310 m, with an inclination of approximately 10°, and a trapezoidal cross-section measuring 8.6 m in width and 3.6 to 4.2 m in height. To facilitate the transportation and installation of fully mechanized mining equipment, three return chambers measuring 5.4 × 3.6 m were excavated within the opening cut. The cross-section of the opening cut support is shown in Figure 1.

3. Research on Similar Model Experiments for Opening Cuts

3.1. Design of Experimental Model

Based on the geological data such as the columnar diagram of surrounding rock properties provided by the Hongliu Coal Mine, calculations and analyses were conducted [43]. A laboratory model was selected for the plane similar material experiment with dimensions of the station as follows: length (1200 mm), width (120 mm), height (1200 mm). The scale ratio of the model is 1:20, with a unit weight ratio of 1:1.6, a time ratio of 1:4.47, and a stress and strength ratio of 1:32. The similar materials comprised river sand, fly ash, gypsum, and calcium carbonate mixed in specified proportions, combined with water, stirred evenly, and then filled into the model frame. For the layered materials, mica powder ranging from 8 to 20 mesh was used, with detailed proportions detailed in

Table 1. Model proportion table.

Number	Rock Type	Stratum Thickness/m	Model Thickness/cm	Proportion Number	River Sand, Gypsum, talc/kg	Layer Thickness/cm	Number of Layers
1	Coarse sandstone	2.5	12.5	728	4.032:0.115:0.461	2	7
2	Fine sandstone	5.8	29	746	4.032:0.230:0.346	2	15
2	Carbonaceous mudstone	0.2	1	837	2.048:0.077:0.179	1	1
3	No. 2 Coal seam	4.8	24		4.608	2	12
4	Siltstone	4.0	20	737	4.032:0.173:0.403	2	10

.

The back of the model is fixed with frosted glass panels, while the front is secured with transparent acrylic glass and includes a reserved opening for roadway excavation. At the top of the model, one hydraulic jack is installed, and below it, two CL-YB-114B-type pressure sensors are placed to monitor changes in the load of the hydraulic cylinders. A steel channel is installed between the oil pump pressurized jack and the top of the model to ensure even distribution of pressure applied by the hydraulic cylinder on the model surface. The specific model setup is illustrated in Figure 2.

3.2. Simulation Design of Opening Cut Support Parameters

The dimensions of the experimental model simulating the opening cut are 430 × 210 mm. Two support cross-sections are to be simulated within the opening cut, with a spacing of 60 mm between them. Transparent epoxy resin-based cementitious material‌ was adopted as the grouting binder for cable and rock bolts in the model. The design of the cross-section support parameters is as follows; for specific support parameter details, refer to Figure 3:

• The roof anchor cable support is designed in a one-beam, three-cable configuration. A 30 mm long fine iron wire is inserted into the roof, with a spacing of 144 mm between the three anchor cables.
• A total of four roof anchor rods are arranged, utilizing 1.5 mm fine iron rods inserted into the roof of the opening cut. Two rods are inserted vertically, while the other two are inserted at a 45° angle at the shoulder of the opening cut. The spacing between the four anchor rods is 144 mm, with a spacing of 72 mm between the roof anchor cables.
• Two side anchor rods, each 12.5 mm in length, are inserted into each side of the opening cut, with a spacing of 69 mm between them.
• The entire roadway is reinforced with steel strips and metal mesh, simulated using iron sheets and fine iron wire mesh, respectively.
• As the pressure on the model increases, when significant deformation of the opening cut is imminent, two elastic single support pillars composed of ordinary pencils and steel pipes, each 210 mm in length, are installed at the two support cross-sections of the opening cut to simulate the on-site single support pillars. The single support pillars are positioned on either side of the centerline of the opening cut, with a spacing of 215 mm between them.
• To facilitate the monitoring of the force changes in the anchor rods (cables), five monitoring points are installed at the top of the left cross-section of the opening cut, consisting of two anchor cables, two roof anchor rods, and one side anchor rod. Strain gauge anchor rods are used for these measurements.

3.3. Monitoring Equipment

The monitoring equipment used for the model and its functions are listed in

Table 2. Monitoring equipment and the function.

Number	Equipment	Function
1	High-precision dial gauge	Monitoring the movement and deformation of surrounding rock in roadway
2	PENTAX R-322NX optical total station	In the model cut, 56 monitoring points are arranged in the overburden rock layers as well as on the left and right sides to describe the displacement changes of the surrounding rock
3	108-channel pressure data acquisition system	Monitoring the stress distribution of the coal pillar in the opening cut
	CL-YB-114B pressure sensor
4	CL-1M-10 static resistance strain gauge	Monitoring the force changes at the ends of strain gauge anchor rod (cable)
5	Circular pressure sensor	Monitoring the stress changes in the single support pillar
	108-channel pressure data acquisition system

. The main monitoring devices are illustrated in Figure 4.

3.4. Scale Limitations and Scale Effect

Although similar material simulation experiments have played an important role in research, the experimental results may have certain limitations due to the limitations of model scaling. Firstly, there is a significant difference in the size of the model compared to the actual situation, which may result in some subtle deformations and damage phenomena that cannot be fully presented in the model. Secondly, although the similar materials and parameters used in the model have been carefully selected, it is still difficult to fully simulate the mechanical behavior of actual rocks. Therefore, when applying experimental results to practical engineering, it is necessary to fully consider the influence of scale effects and make necessary corrections and adjustments. In order to mitigate the impact of scale effects, this study tried to approach the actual situation as closely as possible in the model design and material selection process, and obtained data that were as accurate as possible through various monitoring methods. However, future research still needs to explore more precise methods for designing similar materials and models to improve the reliability and practicality of experimental results.

4. Experimental Procedure

4.1. Excavation and Support of the Opening Cut

The excavation of the opening cut model employs a two-stage expansion method. Initially, a roadway measuring 107.5 × 105 mm is excavated in the central area on the left side of the opening cut’s centerline. This roadway is expanded vertically until it reaches the designed height, and roof support is installed. Next, the roadway is expanded laterally: the left side is extended to the centerline of the roadway, while the right side is expanded to the designed width. During this phase, mesh and anchor rod (cable) support are applied to the roof and right side of the roadway. Subsequently, the right side of the roadway is further expanded by 107.5 mm until it reaches the centerline of the right section of the opening cut, with roof support being installed. The expansion continues to the right until the roadway reaches its designed width, ensuring complete support for the opening cut. After the combined support of the anchor rods (cables) is completed, it is observed that due to the large excavation cross-section, numerous fine cracks have appeared in the overburden rock layers. To maintain the integrity of the surrounding rock, elastic single support pillars are installed for enhanced support, along with the installation of dial gauges. This process is illustrated in Figure 5. With this, the preliminary work for the experiment is complete, allowing for the loading experiments to proceed.

4.2. Model Loading Procedure

The model is subjected to longitudinal loading using a top hydraulic cylinder, with a total of 20 loading cycles, accumulating a load of 15 kN. To ensure the stability of the model under load, the first eight loading cycles increase by 0.5 kN each, while the remaining cycles increase by 1 kN each, as shown in Figure 6.

Data from the instruments are collected before loading and after the loading stabilizes. The data include measurements from the dial gauge, total station, 108-channel pressure sensors, anchor rod load cells, and pressure sensors at the base of the single support pillar.

During the first nine loading cycles, with a cumulative load of 5 kN, it was observed that although a few irregular fine cracks appeared in the overburden rock, and a tendency for roof subsidence, no delamination occurred, and the surrounding rock of the opening cut showed no significant damage.

During the 10th loading cycle, with a cumulative load of 6 kN, damage began to occur at the left side corner of the opening cut, resulting in a crack measuring 3.2 cm in length, oriented at a 60° angle to the roof. Although no delamination was observed in the roof, the number of fine cracks in the overburden rock increased. As the loading progressed, the crack at the side corner of the opening cut gradually widened, and signs of delamination, loosening of both sides, development of cracks, and roof subsidence became apparent. With the increase in loading cycles, the delamination continued to expand, the loosening of the sides increased, cracks continued to develop, and damage occurred at the bottom corners of the sides, accompanied by coal spalling.

During the 12th loading cycle, with a cumulative load of 8 kN, localized failure and minor loosening zones were observed in both sides of the opening cut. The left side exhibited a loosening zone extending 2.4 cm from the opening cut, while the right side showed a 3.4 cm wide loosening zone. Additionally, localized spalling occurred at the right side, accompanied by minor coal fragmentation. A progressive delamination (1 cm thick) was detected in the immediate roof strata, indicating an ongoing subsidence tendency of the roof.

During the 15th loading cycle, with a cumulative load of 11 kN, the cracks on the left and right sides of the incision continued to develop, with a loosening range of 4.4 cm on the left side and 4.1 cm on the right side. The bottom corner of the left side was damaged, with a small amount of coal peeling off and the roof continuing to sink.

By the 19th loading cycle, with a cumulative load of 15 kN, it was observed that cracks on both sides of the opening cut continued to develop. The loosened area on the left side expanded to 5.5 cm, while the right side reached 5.0 cm, with severe surface damage evident on both sides. There was a sudden increase in the amount of displacement, indicating a potential for imminent collapse. The first layer of rock above the roof, located between 22.5 cm and 33.3 cm from the left side, experienced fracturing, leading to a sudden increase in delamination and a risk of roof fall. Additionally, the single support pillar at the first support cross-section on the right side underwent severe bending deformation, ultimately resulting in failure due to compression, as shown in Figure 7. The physical simulation results demonstrated that the roof subsidence (24.071 mm) and sidewall convergence (23.478 mm) in the opening cut exhibited significantly close magnitudes, indicating a synergetic deformation mechanism between the roof and sidewall under mining-induced stresses.

During the 20th loading cycle, while the pressure was expected to reach 16.0 kN, a fracture in the immediate roof occurred at 15.3 kN, rendering the model unable to continue bearing the load. Cracks on both sides of the opening cut developed, with the loosened area on the left side expanding to 7.3 cm and that on the right side to 6.1 cm. The surface damage on both sides was severe, accompanied by the spalling of some coal blocks. A fracture occurred in the first layer of rock above the roof, located between 22.5 cm and 33.3 cm from the left side, with significant delamination and subsidence observed, along with some falling rock blocks. Additionally, two cracks, each measuring 31.4 cm in length and penetrating through the model, appeared on the right side of the right wall. These cracks led to the fracturing of the immediate roof above the opening cut, causing the support to fail and indicating a tendency for overall collapse. During the 20th and final loading cycle, the model suffered complete damage, with severe deformation of the surrounding rock in the opening cut, as shown in Figure 8.

5. Experimental Data Analysis

5.1. Dial Gauge Data Analysis

To monitor the displacement changes of the surrounding rock in the opening cut, a total of five high-precision dial gauges were arranged within the model. Two gauges (Gauge 1 and Gauge 4) were placed on each side of the opening cut, while two gauges (Gauge 3 and Gauge 4) were located beneath the roof. Additionally, one gauge (Gauge 5) was installed on the bottom. These five gauges specifically monitored the convergence of the two sides of the opening cut, the subsidence of the roof, and the bulging of the bottom. The detailed arrangement is illustrated in Figure 9.

After the loading concluded, the cumulative displacement readings from Gauges 1 to 5 were as follows: 14.017 mm, 21.882 mm, 26.260 mm, 9.461 mm, and 0.736 mm, respectively. It is evident that the roof experienced significant subsidence during the compression process, while the sides exhibited varying degrees of damage, with no bulging phenomena observed. Experimental data from dial gauges are shown in

Table 3. Experimental data from dial gauges.

	1#	2#	3#	4#	5#
Initial readings	2.619	3.031	1.068	0.902	5.778
Final cumulative readings	16.636	24.913	27.328	10.363	6.514
Cumulative displacement	14.017	21.882	26.260	9.461	0.736

, and cumulative displacement variation curves of the five dial gauges during loading are shown in Figure 10.

5.1.1. Roof Subsidence Analysis

From Figure 10, it can be observed that during the first 10 loading cycles, the cumulative subsidence of the two dial gauges under the roof was relatively consistent, with both showing minor subsidence. Gauge 2 on the left side recorded a cumulative subsidence of 4.426 mm, while Gauge 3 on the right side measured 4.676 mm. At this stage, the model rock mass was in the initial void compaction phase.

During the 11th to 16th loading cycles, the roof subsidence accelerated significantly, with Gauge 2 recording a cumulative subsidence of 10.695 mm and Gauge 3 measuring 11.610 mm. This increase can be attributed to a larger crack located 33.3 cm from the left side of the roof, which showed signs of potential roof fall. Consequently, the subsidence on the right side of the opening cut was greater than on the left, indicating that the model rock mass was into the elastic deformation phase.

From the 17th to the 20th loading cycles, the roof subsidence increased rapidly. By the end of the experiment, Gauge 2 recorded a cumulative subsidence of 21.882 mm, while Gauge 3 reached 26.260 mm. At this point, numerous cracks appeared in the overburden rock layer within a 4 cm range above the roof, and a fracture occurred in the roof 33.3 cm from the left side, accompanied by rock fall. This observation suggests that the model rock mass had entered the plastic failure phase.

5.1.2. Side Convergence Analysis

From Figure 10, it can be observed that during the first 10 loading cycles, there was almost no displacement in the sides of the opening cut. The left side (Gauge 1) recorded a convergence of 1.222 mm, while the right side (Gauge 4) measured only 0.276 mm. At this stage, there was no significant damage to the rock mass on either side. During the subsequent 10 loading cycles, the convergence rates of the sides accelerated, with the left side experiencing greater convergence than the right side. The changes on the right side were more stable compared to those on the left. This difference can be attributed to the earlier onset of damage at the left shoulder of the opening cut, which led to accelerated deterioration of the rock mass on the left side. Additionally, the loosened area on the left side consistently remained larger than that on the right side. By the 20th loading cycle, two significant cracks that penetrated the model appeared above the right side, resulting in severe deformation of the left side. The convergence for the left side increased by 5.113 mm in a single measurement. At the end of the loading, the total convergence of the roadway was recorded at 23.478 mm, which is comparable to the average roof subsidence of 24.071 mm recorded by the two gauges. It is important to note that the experimental model was not subjected to lateral pressure during testing. Considering that lateral pressure exists in actual field conditions, the convergence of the sides would likely be significantly greater than the roof subsidence, which aligns with on-site monitoring data regarding mining pressure.

5.1.3. Bottom Bulging Analysis

Due to the large cross-section of the opening cut, pressure measurements were not applied to the model, resulting in minimal observable bottom bulging. During the 14th to 17th loading cycles, as the load on the single support pillar increased, the roof experienced compression, leading to a tendency for subsidence. In the 18th to 20th loading cycles, significant damage occurred to the surrounding rock, resulting in a slight upward movement of the roof in specific areas.

5.2. Bottom Pressure Sensor Data Analysis

The model’s base is equipped with 24 pressure sensors, numbered sequentially from right to left as 1, 2, 3, …, 23, and 24. Each sensor measures 5 cm in width, effectively covering the entire bottom of the model. Using the 108-channel pressure data acquisition system, data from the bottom sensors were recorded before and after loading the model. After analyzing the data following 20 loading cycles, pressure variation curves for the sensors were plotted, as shown in Figure 11.

From the above Figure 11, it can be observed that the pressure changes were greatest at Sensor 5, located 37.5 cm from the centerline on the left side, with a variation of 184.16 kPa, and at Sensor 20, located the same distance from the centerline on the right side, with a change of 176.52 kPa, as shown in

Table 4. Main pressure variation values of bottom sensors.

Sensor number	5	8	9	13	16	17	20
Distance from centerline/cm	1.12	0.693	0.663	−2.83	−2.12	−0.56	0.92
Pressure variation/kPa	184.16	50.23	33.01	−7.771	18.28	37.57	176.52

. In contrast, the pressure variation was minimal at Sensor 13, situated at the center, which decreased by only 7.772 kPa. This indicates that the peak pressure support area for the opening cut is located 16 cm from each side, where the surrounding rock is particularly susceptible to failure, potentially leading to spalling of the opening cut. This observation aligns with the significant damage observed at the sides of the model, where the loosened areas reached 9.3 cm and 8.1 cm, respectively. The peak pressure support area, located 16 cm from each side, highlights the regions most susceptible to failure. These findings are crucial for optimizing support designs to target high-stress areas.

To analyze the variation in individual sensors, representative pressure variation values from the right-side Sensors 5, 8, 9, and 13 were selected. The pressure change curves for each loading cycle are plotted in Figure 12.

Analysis of Figure 12 reveals that the pressure on Sensor 5, located 37.5 cm from the centerline, is significantly greater than that of the other three sensors. In contrast, the pressure recorded by Sensor 9, located beneath the opening cut, is noticeably lower than that of Sensor 8, which is situated beneath the left side of the opening cut. The sensor directly under the center of the opening cut experiences the least pressure. During the first 13 loading cycles, the pressure changes for Sensors 5, 8, and 9 exhibit a gradual and linear increase. However, after the 14th loading cycle, the deterioration of the coal walls on both sides of the opening cut accelerates, leading to an increase in the loosened area. As a result, the pressure variations for Sensors 8 and 9 decrease with continued loading, while the pressure variation for Sensor 5 increases. This change occurs because the coal walls near the opening cut become damaged and loose, losing their ability to support the overlying rock layers. Consequently, the rock mass farther from the coal walls starts to take on the main load-bearing role. As loading increases, this further expands the loosened area of the coal walls, ultimately leading to widespread spalling.

5.3. Total Station Data Analysis

A total of 56 measurement points were arranged on the model’s surface. Among these, 44 points were distributed across four rows placed above the opening cut, with each row containing 11 measurement points. The points in the first row are labeled from left to right as a0, a1, a2, …, a10; the second row as b0, b1, b2, …, b10; etc., with the fourth row labeled as d0, d1, d2, …, d10. Additionally, there are two rows of measurement points along both sides of the opening cut, comprising 12 points in total. The first row, from left to right, is labeled e1, e2, e3, e9, e10, e11, and the second row is labeled f0, f1, f2, f9, f10, f11. The total station data collected before loading were plotted as a scatter plot, as shown in Figure 13.

5.3.1. Scatter Plot Analysis of Measurement Points

After each loading cycle, observations were made for all points using the total station. The distribution of measurement points after the experiment was compared to their distribution before loading. Figure 13b indicates that after the 20th loading cycle (with a maximum load of 15 kN), two vertical cracks appeared on the left side of the model, indicating that the model had suffered complete failure. Additionally, significant changes in subsidence were observed at each measurement point compared to the 19th loading cycle.

5.3.2. Analysis of Subsidence Curves for the Overburden Rock Layer

Based on the subsidence data measured from the four rows of points above the opening cut, curves were plotted to represent the subsidence of each row of measurement points as a function of the applied load.

From Figure 14a, it can be observed that when the model was loaded to 12 kN, the subsidence at each point changed slowly. However, after reaching 13 kN, the overall deterioration of the model accelerated, leading to a marked increase in the displacement and subsidence trends of all measurement points. Notably, the subsidence values after the final loading were significantly greater than those recorded during the 19th loading cycle (15 kN). Additionally, the subsidence on the left side of the model was greater than that on the right side, with the measurement points directly above the opening cut showing greater subsidence than those on the sides. Among these, Measurement Point a4 recorded the maximum subsidence of 28 mm, while Measurement Point a10 on the right side showed the minimum subsidence of 9 mm.

Figure 14b presents the subsidence curve for the second row of measurement points, exhibiting a similar trend to that of the first row. Here, Measurement Point b4 also recorded the maximum subsidence of 28 mm, while Measurement Point b10 showed the minimum subsidence of 8 mm. Figure 14c,d indicate that the measurement points in the third and fourth rows follow a similar pattern to those in the second row.

Figure 14e shows the subsidence change curves for four measurement points (a5, b5, c5, d5) along the centerline of the model as a function of loading cycles. It is clear that prior to the 15th loading cycle, the subsidence curves were relatively flat, with minimal changes in subsidence. However, after the 17th loading cycle (13 kN), the curves steepened, indicating significant changes in subsidence, particularly after the 20th loading cycle when the model experienced substantial damage, resulting in greater changes compared to the 19th loading cycle. The displacement curve for Measurement Point d5 consistently remained below the other three curves, indicating that d5 experienced greater subsidence, with delamination reaching up to 2 mm. The curves for the other three points largely overlapped, suggesting that the overburden rock layers remained intact and exhibited overall subsidence under pressure.

5.4. Analysis of Circular Pressure Sensor Data at the Base of the Support Pillar

In this experiment, two circular pressure sensors were installed at the base of the simulated single support pillar in the opening cut. However, the left-side sensor was damaged. Data from the right-side single support pillar, collected through 108 channels, were organized to generate a pressure variation curve, as illustrated in Figure 15.

From Figure 15, it can be observed that during the first 16 loading cycles, the force increment on the single support pillar remained stable, with no visible changes to the exterior of the simulated pillar. However, after the 17th loading cycle (13 kN), the force on the pillar increased significantly, resulting in severe deformation of the single support pillar. By the 19th loading cycle (15 kN), the force reached a peak of 4.0 kN. Following this peak, the pillar experienced a fracture, leading to a rapid decrease in pressure to 2.0 kN as the pillar unloaded. This event also triggered a large-scale collapse of the overburden rock layer above the model. These observations indicate that the single support pillar plays a crucial role in supporting the roof in the opening cut.

Prior to the fracture of the individual prop, severe rib loosening accompanied by localized rib spalling/sloughing was observed, whereas no significant roof subsidence was detected. Following prop failure, the overlying strata exhibited catastrophic structural collapse, resulting in complete loss of load-bearing capacity of the roof system. This critical instability mechanism ultimately triggers a high-risk roof fall hazard due to the abrupt stress redistribution within the surrounding rock mass.

5.5. Uncertainties and Sensitivity Analysis

In the process of analyzing experimental data, we noticed that there is a certain degree of uncertainty in the experimental data, which mainly comes from factors such as experimental conditions, measuring equipment, and human operation. Analysis shows that changes in certain parameters (such as support material strength and model size) have a significant impact on experimental results, while the effects of other parameters (such as environmental temperature and humidity) are relatively small. During the experiment, in order to reduce the uncertainty of experimental data, we also took various measures, including using high-precision measuring equipment, conducting multiple repeated experiments, and strictly controlling experimental conditions. At the same time, in the process of data analysis, we also fully considered the impact of measurement errors and adopted appropriate data processing methods to reduce errors.

6. Discussion

(1)

Before the single support pillar fractures, both sides of the support system experience pronounced loosening and spalling, while the roof exhibits minimal subsidence. After the fracture occurs, the overburden rock layers undergo severe damage and fail to support loads, increasing the risk of toppling accidents.

(2)

Field observation data show that the two sides across the three measurement stations experience 104.858 mm convergence on average, which is noticeably larger than the mean roof subsidence (i.e., 46.000 mm). This indicates that the surrounding rock failure occurs on both sides. In contrast, when the existing support parameters are optimized, the simulated roof subsidence and side convergence are comparable, reaching 24.071 mm and 23.478 mm, respectively, or 0.48 m and 0.47 m in the real-world scenario. This result indicates that the optimized support scheme effectively controls the convergence of the surrounding rock.

(3)

The peak support pressure occurs approximately 16 cm from the coal side, representing 3.2 m in actual conditions. Within the range from the coal side to the peak support pressure, the surrounding rock on both sides is prone to fragmentation and plastic deformation. Through similar material simulation experiments, the optimized loosening and damage range for the left and right sides of the opening cut are 7.3 cm (1.46 m in the field) and 6.1 cm (1.22 m in the field), respectively. Compared to the theoretical calculation of the plastic failure range of 1.83 m under the original design support parameters, the optimized support parameters effectively control the loosening and damage range of both sides of the opening cut.

(4)

Data from three delamination measuring stations indicate that the delamination range of the overburden rock layer above the opening cut is between 0 and 5 m, with almost no delamination occurring between 5 and 8 m, suggesting stability of the roof rock layer at a depth of approximately 5 m. The use of 8.3 m rock cables in the field risks penetrating water-bearing layers, leading to water ingress and softening of the overburden rock, which complicates excavation support. It is recommended to shorten the cable length to 6 m. Observations from the experimental model show that the delamination range of the optimized support parameters is 12.3 cm (2.46 m in the field), indicating that shortening the cable length prevents large-scale loosening of the roof rock layer, making the optimized cable parameters feasible.

(5)

Observations reveal that the average delamination values for the roof rock layers at depths of 1–3 m and 3–5 m are 9 cm and 15 cm, respectively, with the delamination in the 3–5 m range being significantly greater than in the 1–3 m range. This suggests that the top rock rods are effective in controlling fragmentation in the upper rock layers but less effective for deeper rock layers. Through theoretical calculations, it is recommended to increase the length of the top rock rods, with a reasonable length suggested to be 3 m. Theoretical calculations recommend increasing the length of the top rock rods, with a reasonable length suggested to be 3 m. Simulation experiments indicate that the delamination of the optimized roof rock layer is only 12.3 cm (2.46 m in the field), and delamination in the 15–25 cm range (3–5 m in the field) is absent, confirming the feasibility of the top rock rod support scheme.

The on-site monitoring data show good consistency with the simulation experiment results. The optimized support parameters effectively constrained rib convergence and reduced loosening failure zones in open off-cut ribs. Comparative analysis of field monitoring data and simulation results validated the practical applicability of the optimized scheme, demonstrating its effectiveness. Critical control parameters for large-section open off-cut support were identified, enabling rational determination of support configurations including cable bolts and roof bolts.

7. Conclusions

This study systematically analyzes the deformation of large-section open cuts in faces with large mining heights using similar material model experiments. We also characterize the movement patterns of overburden rock layers. In addition, the force acting on the combined support of rock rods, cables, and single support pillars is examined. The findings are summarized as follows:

(1)

In large cross-section openings, deformation and failure initially occur at the side corner of cuts.

(2)

The surrounding rock failure occurs on both sides, and the optimized support scheme effectively controls the convergence of the surrounding rock.

(3)

The optimized support parameters effectively control the loosening and damage range of both sides of the opening cut.

(4)

Shortening the cable length prevents large-scale loosening of the roof rock layer, making the optimized cable parameters feasible.

(5)

Simulation experiments indicate that the top rock bolt support scheme is feasible.

(6)

Integrated analysis of theoretical calculations and field monitoring data yields optimized support parameters: cable bolt length of 6.0 m and roof bolt length of 3.0 m.

Although this study has achieved certain results in optimizing the support parameters of large-section open cuts in high mining faces, there are still some key issues that need further research, which is also an important direction for future research. For example, future research should focus on how to dynamically adjust the support parameters according to the geological conditions and support effects during the mining process, so as to improve the adaptability and flexibility of the support system. At the same time, with the development of artificial intelligence and Internet of Things technology, it is crucial to develop an intelligent support system to achieve real-time monitoring and automatic adjustment of support parameters, which is essential for improving support efficiency and safety.