Optimization of drawing parameters based on top-coal flow law in thick-seam caving mining.

Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Sci Rep . 2026 Mar 4;16:12078. doi: 10.1038/s41598-026-35742-9 Search in PMC Search in PubMed View in NLM Catalog Add to search Optimization of drawing parameters based on top-coal flow law in thick-seam caving mining Shixiong Wu Shixiong Wu 1 College of Mining, Guizhou University, Guiyang, 550025 Guizhou China Find articles by Shixiong Wu 1 , Xun Xu Xun Xu 1 College of Mining, Guizhou University, Guiyang, 550025 Guizhou China Find articles by Xun Xu 1 , Jun Wang Jun Wang 1 College of Mining, Guizhou University, Guiyang, 550025 Guizhou China Find articles by Jun Wang 1, ✉ , Dezhong Kong Dezhong Kong 1 College of Mining, Guizhou University, Guiyang, 550025 Guizhou China Find articles by Dezhong Kong 1, ✉ , Guiyi Wu Guiyi Wu 1 College of Mining, Guizhou University, Guiyang, 550025 Guizhou China Find articles by Guiyi Wu 1 , Qinzhi Liu Qinzhi Liu 1 College of Mining, Guizhou University, Guiyang, 550025 Guizhou China Find articles by Qinzhi Liu 1 , Yujun Zuo Yujun Zuo 1 College of Mining, Guizhou University, Guiyang, 550025 Guizhou China Find articles by Yujun Zuo 1 Author information Article notes Copyright and License information 1 College of Mining, Guizhou University, Guiyang, 550025 Guizhou China ✉ Corresponding author. Received 2025 Nov 8; Accepted 2026 Jan 7; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13077052  PMID: 41781407 Abstract With the continuous advancement of coal mining technology and widespread application of modern equipment, fully mechanized top-coal caving (FMTC) mining has become an efficient and economical option for thick coal seam mining. To study its coal drawing process, this paper takes the FMTC mining at Working Face 11508 of Xiejiagou Coal Mine in Guizhou as the research background. Via numerical simulation and bulk similar simulation experiments, it explores the law of top-coal drawing, recovery rate and gangue content under different drawing processes at the working face, clarifies the "window-closing" principle for drawing control, and optimizes drawing process parameters. Results show: When the mining-drawing ratio is 1:1, top-coal output is most balanced across steps; when it is 3:1, output difference is most obvious. The best top-coal drawing effect is achieved with a drawing step of 0.8 m and mining-drawing ratio 1:1.5 (1.96 m mined, 2.94 m drawn). At 1:1.5 ratio with 1-mining-1-drawing, though gangue content is higher, top-coal drawing rate is the highest and periodic output is stable. Strictly implementing the "close window when gangue is seen" termination principle in practice can achieve high top-coal recovery while controlling gangue content, providing guidance for thick seam FMTC mining. Keywords: Coal seam thickness; Completely mechanized mining in caves, and optimization of mining proportions of coal tunneling step distance optimization; Coal caving sequence optimization Subject terms: Energy science and technology, Engineering Introduction Top-coal caving (TCC) has emerged as a landmark technological achievement in China’s coal mining industry, exerting significant international influence 1 – 3 . With the gradual depletion of easily minable coal reserves, the exploitation of deep thick coal seams has become a core priority for the Chinese coal industry 4 – 7 . However, the high-intensity underground mining of such seams is frequently associated with problems like roof collapse and coal wall spalling, substantially increasing the technical difficulty 8 – 12 . China possesses abundant thick coal seam reserves. The efficiency of their extraction not only directly determines the overall coal production but also critically influences the coal quality. Compared to conventional fully mechanized mining, the fully mechanized top-coal caving (FMTC) method involves greater complexity, with key operational parameters including the caving-to-mining ratio, caving interval, and caving sequence 14 – 21 . Extensive research has been conducted on the key technical issues in FMTC mining of thick coal seams. Wang et al. 22 elucidated the top-coal collapse mechanism through physical simulations and numerical modeling of field cases, proposing a high-accuracy marking method for calculating the top-coal recovery rate. Liu et al. 23 employed numerical simulation to specifically investigate the recovery rate of a 4.0 m top-coal layer, providing practical guidance for selecting caving intervals in thick seams. Zhao 24 and Zhang et al. 25 . applied the Continuous-Discontinuous Element Method (CDEM) to simulate the collapse and flow of top-coal fragments, offering technical support for automated and efficient TCC in extra-thick seams. Tien Dung Le et al. 26 , 27 proposed a discontinuous modeling approach, revealing that top-coal failure is shear-dominated while overburden failure is tension-dominated, thereby clarifying the core factors influencing top-coal fragmentation. Yang et al. 28 innovatively proposed the Independent Cluster Group Caving Technology (ICGCT), demonstrating its effectiveness in increasing top-coal recovery at the upper section of the working face. Zhang et al. 29 utilized top-coal pre-blasting to weaken the coal strength, improving caving characteristics, and obtained recovery rate statistics via numerical simulation, contributing to coal loss reduction. Zhang, Song et al. 30 , 31 conducted simulation and back-analysis of top-coal caving using the 2D discrete element software PFC, providing a theoretical basis for improving recovery rates in fully mechanized mining of ultra-thick coal seams. Maximizing the top-coal recovery rate is the central goal of TCC mining. While existing studies have covered mechanisms, simulations, and process optimization, systematic research on the refined optimization of caving parameters and the definitive "cut-off criterion" for caving control under specific geological conditions remains insufficient. To address this gap, this study focuses on the 11508 FMTC working face at the Xiejiagou Coal Mine in Guizhou Province. By integrating numerical simulation with granular similarity simulation experiments, we systematically investigate the top-coal drawing patterns, recovery rates, and gangue content characteristics under different caving parameters. The objective is to establish a clear cut-off criterion for caving control and ultimately optimize the caving parameters, providing a technical reference for FMTC mining under similar geological conditions. The general situation of 11508 at Xiejiahegou Coal Mine, the caving face is completely automated The No. 15 coal seam is one of the primary mineable seams in Xiejiahegou Coal Mine. The main seam being mined is the No. 15 coal seam, which is the middle section of the Longtan Formation in the upper part of the 11508 fully automated top-coal caving face. The seam is stable and recoverable throughout the entire area. The upper distance from the top boundary of P 3 l 2 is 8.77–30.96 m, averaging 17.56 m. The minimum inter-seam distance to the No. 16 coal seam ranges from 1.61 to 8.12 m, with an average of 5.35 m. The thickness of Coal Seam No. 15 ranges from 1.29 m to 11.43 m, with an average of 4.9 m, making it a thick coal seam. The coal seam’s dip angle is 16°–22°, which is classified as a gently dipping coal seam.. The coal seam contains one to two localized partings (gangue layers) in addition to the outer layer. The overall structure is relatively simple, and the roof and floor stability is good, indicating a stable coal seam.The lithology of the roof and floor of the No. 15 coal seam is detailed in Table 1 . Table 1. Lithology of the roof and floor of the No. 15 coal seam. Roof and floor Rock name Thickness (m) Lithological characteristics Main roof Siltstone 7.31–14.18 Gray, hard, sliding surface is obvious, local broken and see vertical cracks, easy to weathering Average 9.13 Immediate roof Muddy siltstone 4.93–13.17 Light gray–dark gray, thin-medium thick layered, wavy bedding, fissure development, calcite filling, rich in plants, compressive strength is larger, more stable Average 8.43 Coal seam No. 15 coal 1.29–11.43 Black, powder-based, interbedded block, bright coal-based, glass luster, semi-bright, generally 0–1 layer of stone, simple structure, is a more stable coal seam Average 4.9 Direct bottom Mudstone 0.2–2.8 Gray to dark gray, thin-medium thick layer, wavy bedding, fracture development, containing fine sand and argillaceous bands, mud in water, easy expansion, bottom drum phenomenon Average 2.5 Basic bottom Fine Sandstone 6.2–9.8 Light gray to dark gray, thin-medium thick layer, wavy bedding, fracture development, mainly sandstone, rich in plant Average 8.2 Open in a new tab Mechanical analysis of completely automated caving face Coal rock is tested mechanically and physically in working face Through the mechanical experiment of the standard specimens made of 11508 working the mechanical characteristics of rock and coal samples the operating face’s coal seam roof (11508) are determined, as well as the Rock mass and coal mechanical characteristics are determined, which provides a reference for studying top coal and the top coal cavern caving law shows a caving face that is completely automated. The roof rock block of 11508 working face was processed into cylinder samples with diameters of 50 × 100 mm, 50 × fifty millimeters, and 50 × 25 mm the sizes, and the two ends of the time were smoothed by the mill. The uniaxial compression test were conducted using a servo-controlled hydraulic testing machine.. The test data was recorded by a computer system with a sampling frequency that can reach 0.1 ms. The experimental equipment and rock specimens are shown in Fig. 1 . Fig. 1. Open in a new tab Hydraulic material testing machine and fixture. Based on uniaxial mechanical tests and mine geological data, the mechanical parameters of the coal-rock mass for Working Face 11508 are presented in Table 2 . Table 2. Experimental test Findingson the Rock’s mechanical and physical properties mass and coal from the No. 15 coal seam. Lithologic characters No Compressive strength/MPa Mean value Tensile strength/MPa Mean value Cohesion/MPa Internal friction angle/° Immediate roof Siltstone 1–1 55.59 53.9 12.82 13.17 32.7 6.314 1–2 55.32 12.25 1–3 52.79 14.45 Main roof Muddy siltstone 2–1 17.05 15.28 3.52 3.71 29.57 4.891 2–2 13.47 4.38 2–3 15.33 3.25 Coal seam 3–1 8.49 7.09 1.03 0.45 1.57 28.74 3–2 7.21 0.86 3–3 5.57 0.26 Open in a new tab The functional relationship between the principal stresses satisfies: 1 Bring into ( 1 ): 2 where in the formula : —The maximum main stress (Mpa); —The minimum principal stress (MPa); Rc—Top coal’s uniaxial compressive strength, measured in MPa; f—Internal friction coefficient, wherein φ is the frictional angle and f = tanφ; φ = 28.74°; K—Advanced abutment stress concentration factor; H—Mining depth. Substituting the occurrence condition parameters of 11508 working face γ = 25 kN/m 3 (average unit weight of overlying strata), μ = 0.31 (Poisson’s ratio) ,and K = 2 (stress concentration factor) into Eq. ( 2 ), a calculated mining depth of H = 213 m is obtained. The actual average burial depth of the No. 15 coal seam in Xiejiahegou Coal Mine is approximately 425 m, Since the actual burial depth of the coal seam is greater than the critical mining depth, the mine pressure conditions are favorable for top coal crushing, and the extraction depth meets the requirements of relevant regulations. As shown in Fig. 2 , the uniaxial compressive strength, bond coefficient, and internal friction angle are key parameters used to quantify a coal seam’s resistance to failure and deformation. These three properties collectively define the overall strength of the coal seam. Based on the Mohr–Coulomb Failure Criterion, the functional relationship between the three is as follows : 3 4 where: —Uniaxial compressive strength of the top coal (Mpa); C—Cohesion (Mpa). Fig.2. Open in a new tab Top coal body stress diagram. Substituting Eqs. ( 3 ) and ( 4 ), we obtain: 5 Derivation of formula ( 3 – 5 ): 6 Given that the cohesion C is greater than zero (C 0),we can conclude that: . When φ is in the range of (0, π/2 ), combined with Eq. ( 6 ) and Fig. 3 , it can be seen that when the cohesion is constant, There is a favorable correlation between the angle of internal friction and coal’s uniaxial compressive strength, so to a certain extent, the uniaxial compressive strength the shape The structure of a coal seam has a substantial impact on coal’s resistance to failure and deformation, so its uniaxial motion compressive strength of coal seam can be used as an index to represent the strength and stress response ability of coal seam. Fig. 3. Open in a new tab Relationship of mechanical parameters of top coal. The caving angle Moreover, premium coal has a very important indirect impact on the uniaxial compressive strength out of coal. According to engineering practice, it is shown that the top coal sinking degree has a negative relationship with the coal seam’s one-way compressive force. Reducing the caving angle of upper coal will make it harder to release the top coal, or worse caving. Figure 4 illustrates the connection between the top coal’s collapsing tendency toward the coal seam underneath it hardness coefficient (f = Rc/10). Fig. 4. Open in a new tab Connection from the caving angle of the top coal to the coal seam’s hardness coefficient. Based on the method for calculating an acceptable caving angle for top coal : 7 where inside area formula : α—Reasonable caving angle; H—Coal thickness, with a value of 11.43 m ; Hg—Coal cutting height, 3.2 m; L—Width fromthe support top beam’s rear end to the inner border of the rear scraper, 3.0 m. Substituting the above values, it can be obtained that when the mining thickness 11.43 m is the fully automated caving face’s appropriate caving angle top coal is 70°. The uniaxial compressive strength (UCS) of the No. 15 coal seam in Xiejiahegou Coal Mine is taken as 7.09Mpa on average, with a hardness coefficient f of 0.7. As shown in Fig. 4 , the measured caving angle of the No. 15 top coal is approximately 110°. When the fully mechanized caving face is mined at a thickness of 11.43 m, the actual caving angle of the top coal is greater than the critical caving angle required for good caving performance. Therefore, the caving efficiency is considered favorable. Overview of working face loose circle The rock mass contains geological weak planes and structures, which seriously reduces the overall strength of the rock mass and increases its deformability. The application of the fully mechanized top-coal caving technique in domestic mining areas demonstrates that the coal’s joints, bedding planes, and fissures have the greatest influence on the stability and performance of the caving face. The degree to which coal seam joints and cracks have developed is negatively correlated with the integrity and general strength of bituminous body. Overall, the strength of the coal body is low, and its integrity is poor. Therefore, the action the abutment stress readily breaks top coal. Furthermore, the fragmentation and caving of top coal decrease with increasing fissure density. That is, higher fissure density leads to better caving performance. Therefore, a higher degree of cavability is beneficial. According to the elastic–plastic hypothesis, the loose circle is the surrounding rock zone where the stress redistributes after roadway excavation and the rock mass is fractured and unconsolidated. It is known as the notion of surrounding rock loose circular support, because it attaches importance to its broken expansion force in the formation process, its application in the classification of surrounding rock, its relationship with the mechanism of bolt action and the corresponding support technology. Although the loose circle’s size can be correlated through model tests, field test and theoretical analysis, the current application in the project is influenced by several elements, including rock properties, ground stress, structure and mining influence. Furthermore, calculating the radius of the loose zone based purely on theoretical formulas is often not recommended due to inherent simplifications and multiple assumptions. Therefore, the borehole TV digital imaging technology is used to observe and measure the rock around the borehole and calculate the loose circle’s radius around the road’s surrounding rock. The Xiejiagou Coal Mine borehole peeper spotted the loose circle of 11508 completely automated caving faces, and the development degree of joint fissures in coal seam was judged, as shown in Fig. 5 . Fig. 5. Open in a new tab TYGD10 rock drilling peeper From Fig. 5 , A coal crack close to the top coal is visible. From the borehole orifice to a depth of 1.0 m in the 11508 open-off cut, an annular fracture zone is observed. At a drilling depth of 2.0 m, the number of cracks significantly decreases. It can be seen that the borehole not only has annular cracks, but also shows a certain deformation. When the drilling depth reaches 3.0 m, a small number of longitudinal cracks appear, accompanied by transverse cracks to form cross cracks. When the drilling depth reaches 4.0 m, the integrity of the coal body appears relatively high; although obvious longitudinal cracks are present, the overall crack density decreases. When the drilling depth reaches 5.0 m, the degree of coal fragmentation is more intensified, and there is a more obvious ring-banded interval fracture zone ; When the drilling depth reaches 6.0 m, the fractures are primarily transverse and annular, with the surrounding rock being mildly fractured. It is precisely Because of these fissures, the top coal may eventually collapse, making it convenient for the top coal to be released. To accurately assess the influence of top coal joint fissures on cavability, the fractal dimension of the joint fissure size distribution is utilized for quantitative evaluation. 8 where in the formula: n(x₀)—Number of joint fissures with a length greater than L₀ within the x₀ × x₀ measurement range; x₀—Fissure length; a₀—Proportional coefficient; D—Fractal dimension of the length-number distribution of joint fissures. The relationship between the uniaxial compressive strength and the fragmentation fractal dimension of coal is: 9 Rc—Uniaxial Compressive Strength. The number of through fissures within the range of any region x E × x E, denoted as n ( x E), is: 10 where in the formula: . The number of penetrating cracks in the measurement range of 1 m × 1 m is recorded as n(1), and the relationship between it and uniaxial compressive strength Rc is : 11 Therefore, the product Dn(1) of D and n(1) is used. Figure 6 shows the relationship curve between this product and the maximum coal digging rate of the working face. It serves as a classification index for the influence of roof joint fissures on roof caving. Fig. 6. Open in a new tab Peep view of top coal borehole at 11508 open-off cut position. Figure 6 shows that the caving condition of the top coal improves as the Dn(1) index increases.When Dn(1) is less than 4, the top coal’s recovery rate is below 50%, indicating poor caving characteristics. When 4 < Dn(1) < 20, the top coal’s recovery rate increases gradually, but the growth rate slows down, indicating that its influence on the caving behavior gradually diminishes. When Dn(1) is greater than 20, the growth rate of the recovery rate essentially remains constant, indicating that the further increase in Dn(1) has minimal additional impact on the caving characteristics of the top coal. Combining the mechanical parameters of the No. 15 coal seam at the Xiejiahegou Coal Mine, and solving simultaneous Eqs. ( 9 ) and ( 11 ), yields Dn(1) = 15.65. Since 15.65 falls within the optimal range (4 < D n(1) < 20), we conclude that the top coal in the No. 15 coal seam at the Xiejiahegou Coal Mine has well-developed fissures, which significantly helps to improve the top coal’s caving property. Numerical simulation of process parameter optimization Creation of a numerical model The advancement of a fully-mechanized caving face is a continuous process. The coal-rock interface terminated by gangue exposure during the preceding caving operation serves as the initial coal-rock interface for the subsequent caving operation, and significantly influences it. Therefore, it is highly necessary to investigate the state of coal recovery in the top-coal drawing stage.Combining the parameters of the 11508 working face deposition conditions from the preceding text with the data in Table 3 , a model was established. Along the dip, the model length is 30 m and the height is 11.93 m, with a top-coal thickness of 3.5 m and a immediate roof thickness of 8.43 m. A total of 20 supports were set up, with 5 supports retained at each side . Along the strike, a model with a length of 36 m and a height of 13.33 m was established, where the coal seam thickness is 4.9 m and the immediate roof thickness is 8.43 m. 10 m coal pillars were retained at both ends of the model, and the longwall top-coal caving face advanced 20 times.The upper boundary of the model is a free boundary, and the two ends are fixed boundaries. In the model, the longwall top-coal caving support was constructed using the ‘wall’ command, and the movement and coal drawing operations were implemented by writing fish language. The ‘gangue exposure shut-off’ principle was also realized through a custom fish function. All particles within the model have initial displacement and velocity of zero, and are only subjected to gravity (g = 9.8\m/s 2 ) before the initial coal drawing. The model is considered to be in equilibrium when the maximum unbalanced force ratio inside the model reaches e −5 . Table 3. Coal rock particles’ mechanical and physical characteristics. Rock formation Thickness /m Density ρ/(kg/m 3 ) Grain size /mm Normal stiffness kn/(N/m) Tangential stiffness ks/(N/m) Friction factor Porosity Coal seam 4.9 1440 100–200 2 × 108 2 × 108 0.4 0.12 Immediate roof 8.43 2500 100–200 4 × 108 4 × 108 0.4 0.12 Open in a new tab In the numerical model, the simulation experiment is not carried out in front of the support. The black part is regarded as the broken coal seam of the granular body, and the gray part is regarded as the broken immediate roof of the granular body. The color marker line is formed by dyeing the particles in a certain layer to observe the coal body’s properties and flow condition. Table 3 displays the mechanical and physical features of the coal seams and the nearby roof particles, whereas Fig. 7 displays the model’s starting condition. Fig. 7. Open in a new tab Curve diagram of connection between the caving rate of top coal and fracture distribution characteristics. Due to the gaps and voids between individual particles, the coal-rock interface is non-smooth, exhibiting undulations (concave or convex features). Therefore, the coordinates of contacting coal and rock particles within a specific range were iterated through using fish language. As shown in Fig. 8 and Fig. 9 , regression analysis revealed that both the initial coal-rock interface and the final coal-rock interface at the termination of initial coal drawing can be fitted by a parabolic curve, thus demonstrating the rationality of the fitting parameters. The fitting results for the dynamic evolution of the coal-rock interface using the fitted curves are as follows: Fig. 8. Open in a new tab The initial state of the model. Fig. 9. Open in a new tab Parabolic Fitting of the Initial Coal-Rock Interface. Initial Coal-Rock Interface (Blue): Final Coal-Rock Interface (Red): Final Coal-Rock Interface at the Termination of the Initial Coal Drawing (Blue): Final Coal-Rock Interface at the Termination of the First Support Advancement and Coal Drawing (Red): Optimization of caving step distance Utilizing the production parameters of the Xiejiahegou 11508 Longwall Top Coal Caving (LTCC) face, a numerical model advancing along the coal seam strike was established. The fixed mining height was set at 1.5 m with a total seam thickness of 4.9 m and different Drawing Step Distances: 0.8 m, 1.6 m, and 2.4 m. Due to the fact that the top coal occurrence conditions of each numerical model are the same under different coal caving step distances. The final coal release quantity and the coal-rock interface morphology are the same after the first coal drawing. Therefore, only the impact of a single primary coal drawing on the entire coal caving process is investigated at various coal drawing step distances. The top coal recovery situation is presented in Fig. 10 . Fig. 10. Open in a new tab Parabolic Fitting of the Final Coal-Rock Interface at the Termination of the Initial Coal Drawing. If the effective diagonal width of the coal drawing port is less than the coal drawing step distance, a sizable portion of the top coal beyond the shield frame of the support falls directly on to the goaf surface after the support is moved, according to the established coal loss mechanism. Some of the top coal on one side in the goaf gets disturbed by the ongoing coal caving and subsequently flows into the coal drawing port. As coal drawing proceeds cyclically, the recovered top-coal body forms an uneven strip inclined toward the goaf side, as shown in Fig. 10 . The width of the strip increases with the coal caving step distance. The number of coal particles, the size of the top coal particles, and the proportion of recovered top coal were all determined using the fish language. The top-coal recovery situation for the fully-mechanized caving face under different caving step distances was statistically analyzed. Due to the application of the Gangue Exposure Shut-off (GESCO) criterion in the simulation, it is impossible to statistically count the release of immediate roof particles. Table 4 presents the statistical findings. The standard deviation was calculated based on the results of multiple coal draws, and thus does not reflect the difference between a single draw and multiple draws. Table 4. Top coal caving statistics of different caving step distance. Mining-to-caving ratio Total recovered top coal (kg) Top coal recovery rate (%) Avg. top coal per cycle (kg) Total caving cycles (Draws) Standard deviation 1:1 17,467 81.25 583 8 387 2:1 16,571 78.63 1696 3 640 3:1 13,843 72.89 1064 4 817 Open in a new tab The calculation method for the standard deviation is as follows: 1. Determination of the Data Series (X): For each mining-to-caving ratio (e.g., 1:1), the data series X refers to the set of all single coal draw quantities under that specific ratio. 2. Calculate the arithmetic mean (i.e., Avg. Top Coal per Cycle in the table): 12 3.Calculate the standard deviation: 13 where: X i —The actual amount of coal caved during the i caving cycle. u—The arithmetic mean coal caving amount (Avg. Top Coal per Cycle) for that mining ratio. n—The total number of caving cycles for that mining ratio. Under the 1:1 mining-to-caving ratio (one mining, one caving): The total recovered top coal was 17,467 kg, resulting in a top coal recovery rate of 81.25%. This strategy involved a total of 8 caving cycles (draws), with the average top coal recovered during each periodic caving operation being 583 kg. Under the 2:1 mining-to-caving ratio (two mining, one caving): The total quantity of caved top coal reached 16,571 kg, meaning 78.63% of the available top coal was extracted. This occurred over a total of 3 top coal caving cycles, where the mean amount of top coal excavated during the periodic caving period was 1696 kg. Under the 3:1 mining-to-caving ratio (three mining, one caving): The total recovered top coal quantity was13,843 kg, yielding the lowest top coal recovery rate at 72.89%. This strategy involved a total of 4 top coal caving cycles overall, with an average of 1064 kg caved over the periodic caving period. The comparison reveals that, ranked from high to low, the top coal recovery rate under various mining-to-caving ratios is: 1:1 mining-to-caving ratio > 2:1 mining-to-caving ratio > 3:1 mining-to-caving ratio. Furthermore, the variance in the recovered coal quantity throughout the periodic caving period demonstrates that the top coal release is distributed most uniformly among each caving step when the extraction and caving combination is 1:1. Conversely, the most pronounced difference in the released top coal between each step occurs when the 3:1 ratio is employed. In conclusion, the 1:1 mining-to-caving strategy provides the most significant positive impact on top coal recovery, yielding the highest recovery rate. Mining ratio optimization In addition to the actual output of the completely automated caving face of Xiejiahegou 11508, the top collie drawing law is examined under the mining ratios of 1: 1, 1: 1.5, 1: 2, and 1: 3, respectively, and a numerical model of progressing down. The natural coal vein is built. We set the thickness of the seam of coal at 4.9 m, and the associated mining heights are 2.45, 1.96, 1.63, and 1.23 m, respectively. One mine and one caving, with a step distance of 0.8 m for coal caving are selected, and a total of 20 steps are advanced. The initial model of numerical simulation is shown in Fig. 11 . Fig. 11. Open in a new tab Various caving step lengths in a top coal caving scenario. Initial coal caving and inversion results Figure 12 shows the interface between the uppermost top coal drawing body and the coal-rock contact surface under various mining-to-caving ratios in fully mechanized top coal caving.. When the mining-to-caving ratios are 1:1 and 1:1.5, the Fig indicates that the top coal release area is larger, the top coal flow velocity is faster, and the coal-rock interface between the support structure and the goaf remains relatively gentle throughout the drawing process. Conversely, when the mining-to-caving ratio ranges from 1:2 to 1:3, there is a noticeable concavity of the coal-rock interface on the goaf side, and the interface becomes steeper on the support side. This results from a reduction in the highest coals release room, an increase of the top coal thickness, and a slow top coal flow. Fig. 12. Open in a new tab Initial model of numerical calculation. An analysis of the inverted top coal caving geometry reveals the geometry of the upper coal drawing opening. When the mining and caving ratios are 1:2 and 1–3, the horizontal width is wider and both extend far over the support shield beam’s edge. Furthermore, the upper edge of the opening is occupied by falling immediate roof gangue. The gangue flows toward the vicinity of the coal drawing port during the initial movement. At the beginning of the subsequent caving, the gangue can easily enter the coal drawing aperture, which reduces the top coal caving rate and may prematurely terminate the caving process before the top coal is fully released. When the mining or caving ratios are 1: 1 and 1: 1.5, the right edge of the opening body is close to the edge of the shield beam, allowing top coal to be discharged. Therefore, the 1: 1 and 1: 1.5 mining & caving ratios are better than the other two from the standpoint that the initial coal caving aids the future top coal caving. The whole process of coal caving and inversion results Figure 13 shows the top coal release situation and inversion diagram of 20 advance steps under different mining and caving ratio conditions. The different color particles in the Fig represent the top coal release particlesat various advance distances. The Fig shows that after the initial caving cycle, the top coal drawing volume is minimal and is significantly affected by the first coal discharge. The drawing particle aggregate during the periodic top coal extraction phase is depicted as an irregular strip inclined towards the goaf side. The length of this drawing strip increases as the mining-to-caving ratio decreases, but the volume of top coal drawn during this periodic phase is disproportionate. Fig. 13. Open in a new tab Initial coal drawing situation. The mass of the drawn coal particles, along with the number and diameter of the top coal particles, are quantified using digital image analysis, and the top coal drawing volume after 20 advance steps of a fully mechanized caving face under various mining-to-caving ratios is calculated. Because a gangue-sensing and door-locking control strategy is implemented in the simulation experiment, it is not possible to count the release of immediate roof gangue particles. The statistical results appear in Table 5 . The standard deviation was calculated based on the results of multiple coal draws, and thus does not reflect the difference between a single draw and multiple draws. The standard deviation is calculated using Formulas 12 and 13 . Table 5. Presents the top coal drawing statistics under various mining-to-caving ratios. Mining-to-caving ratio Total recovered top coal (kg) Top coal recovery rate (%) Mean periodic mass (kg) Cycles > Mean standard deviation 1:1 15,431 84.61 449 9 394 1:1.5 16,021 85.92 526 9 293 1:2 16,571 84.97 540 8 361 1:3 18,953 93.20 664 8 556 Open in a new tab For a mining-to-caving ratio of 1:1: The total recovered mass of top coal is 15,431 kg, resulting in a recovery rate of 84.61%. The mean recovered mass per periodic caving cycle is 449 kg, with a total of 9 caving cycles exceeding this mean. For a mining-to-caving ratio of 1:1.5: The total recovered mass of top coal is 16,021 kg, with a top coal recovery rate of 85.92%. The mean recovered mass per periodic caving cycle is 526 kg, with a total of 9 caving cycles exceeding this mean. For a mining-to-caving ratio of 1:2: The total recovered mass of top coal is 16,571 kg, with a top coal recovery rate of 84.97%. The mean recovered mass per periodic caving cycle is 540 kg, with a total of 8 caving cycles exceeding this mean. For a mining-to-caving ratio of 1:3: The total recovered mass of top coal is 18,953 kg, with a top coal recovery rate of 83.2%. The mean recovered mass per periodic caving cycle is 664 kg, with a total of 8 caving cycles exceeding this mean. In summary, a mining to cave ratio of 1: 1.5, a caving step length of 0.8 m, and an average coal seam width of 4.9 m in the 11508 working face result in the largest coal caving effect, that is, the mining is 1.96 m and the caving is 2.94 m. The top coal recovery rates under different mining-to-caving ratios, ranked from highest to lowest, are: 1:1.5 > 1:2 > 1:1 > 1:3.Drawing Consistency: Analysis of the standard deviation of the recovered mass during periodic caving indicates that the consistency of the top coal drawing volume per step is highest when the mining-to-caving ratio is 1:1.5. Conversely, the discrepancy in the top coal drawing volume per step is most pronounced when the mining-to-caving ratio is 1:3. Comprehensive optimization analysis of process parameters based on orthogonal experiment To systematically investigate the combined effects of the two key operational parameters—the ratio of mining to drawing and the drawing interval—on the top-coal recovery efficiency, this section adopts the orthogonal experimental design method. An L12(4 1 × 3 1 ) mixed-level orthogonal array was constructed, incorporating four levels of the mining-to-drawing ratio (1, 0.67, 0.5, 0.33) and three levels of the drawing interval (0.8 m, 1.6 m, 2.4 m). With the top-coal recovery rate (%) as the core evaluation indicator, twelve sets of numerical simulation experiments were conducted. The orthogonal experimental design is presented in Table 6 . Table 6. Orthogonal experimental design. No Mining-to-drawing Ratio Drawing Interval (m) Experiment (Recovery Rate, %) 1 1 0.8 84.61 2 1 1.6 78.63 3 1 2.4 72.89 4 1/1.5 0.8 85.92 5 1/1.5 1.6 79.83 6 1/1.5 2.4 74.02 7 1/2 0.8 84.97 8 1/2 1.6 78.97 9 1/2 2.4 73.21 10 1/3 0.8 83.2 11 1/3 1.6 77.31 12 1/3 2.4 71.68 Open in a new tab Through range analysis and variance analysis (Table 7 ), the significance of the two factors’ influence on the recovery rate was systematically evaluated. The results show that the range R for the drawing interval is 11.72%, significantly higher than the range of 2.53% for the mining-to-drawing ratio. The corresponding adjusted ranges R’ (accounting for the influence of different level counts) are 12.19 (drawing interval) and 1.97 (mining-to-drawing ratio) respectively, further confirming that the drawing interval is the dominant factor affecting the top-coal recovery rate, with its influence intensity approximately 6.2 times that of the mining-to-drawing ratio. Table 7. Results of range analysis and analysis of variance for the orthogonal experiment. Factor Number of levels Replicates per level Conversion coefficient d Range R (%) Adjusted range R’ Variance of level means Mining-to-drawing ratio 4 3 0.45 2.53 1.97 0.82 Drawing interval (m) 3 4 0.52 11.72 12.19 22.92 Open in a new tab As shown in Fig. 14 , the analysis of the mean values for each factor level indicates that as the drawing interval increases from 0.8 to 2.4 m, the average recovery rate monotonically decreases from 84.68 to 72.95%, representing a significant drop of 11.73 percentage points. This demonstrates that an increase in the drawing interval has a pronounced negative impact on the recovery rate. In contrast, the influence of the mining-to-drawing ratio on the recovery rate exhibits a non-monotonic trend of first increasing and then decreasing: when the ratio is optimized from 1 to 0.67, the average recovery rate rises from 78.71% to a peak of 79.92%; however, as the ratio further decreases to 0.33, the recovery rate instead declines to 77.40%. This suggests the existence of an optimal range for the mining-to-drawing ratio (approximately 0.67), which can effectively balance the drawing space for top coal against the risk of waste rock mixing, thereby maximizing the recovery rate. Fig. 14. Open in a new tab The whole procedure for extracting coal. Therefore, based on the orthogonal analysis, the optimal combination of process parameters is determined to be a drawing interval of 0.8 m and a mining-to-drawing ratio of 0.67. This combination achieved the highest recovery rate of 85.92% in the simulations, corresponding to a total drawn top-coal mass of 16,021 kg, along with the smallest standard deviation of periodic drawing (293), indicating the most balanced and stable drawing process. Under the coal seam conditions of the No. 11508 working face, adopting a process combination of a small drawing interval (0.8 m) paired with a moderate mining-to-drawing ratio of approximately 1:1.5 (~ 0.67) can maximize the reduction of top-coal loss, suppress waste rock mixing, and represent the optimal technical solution for achieving both high recovery rates and balanced drawing. Similar simulation study on process parameter optimization Experimental platform and similar materials The custom-built physical simulation platform is used to simulate the top coal drawing process of fully mechanized top coal caving. The dimensions of the experimental setup are 120 cm × 42 cm × 150 cm (Length × Width × Height). Along the path of the completely automated caving face arrangement, a total of six moveable iron sheets are placed to simulate the coal drawing port, which are numbered sequentially from No. 1 to No. 6. The model support components, drawing ports, and the model roof and floor are constructed from stainless steel plates, and the experimental platform is shown in Fig. 15 . Fig. 15. Open in a new tab Trends of the Mean Effects of Mining-to-Drawing Ratio and Drawing Interval on Recovery Rate. The top coal and the immediate roof are in a highly fractured state throughout the fully mechanized top coal caving process, with significant damage occurring even prior to caving. The stress from the main roof is concentrated entirely on the coal wall, while the residual stress in the goaf is granular, resulting in a compacted zone in the goaf. This phenomenon is typically caused by a large caving step distance. In the experiment, it is not simulated as granular particles. Therefore, in the similar simulation experiment, only two layers of particles with different particle sizes are laid to simulate the direct roof and coal seam. In the experiment, the coal seam is simulated by black water-washed gravel with a particle diameter of 8–12 mm, and the immediate roof is simulated by white water-washed gravel with a particle diameter of 15–25 mm. The geometric similarity ratio is 1:30, as shown in Fig. 16 . Fig. 16. Open in a new tab Similar simulation experiment platform. Based on the field conditions and the optimal top coal thickness determined from numerical simulation, the immediate roof height is set to 8.43 m, and the caving height is selected to be 3 m. In the model, a 10 cm-thick layer of black water-washed gravel is laid to simulate the top coal, followed by a 28.1 cm-thick layer of white water-washed gravel simulating the immediate roof. No coal seam is built in front of the model support. The advancement of the caving face and the movement of the support are controlled by the universal wheel mechanism at the bottom. To simulate the caving cycle: The coal drawing port is opened to discharge the simulated top coal. The port is then closed once the immediate roof particles enter the aperture and contact the model roof, thereby completing the drawing cycle, as shown in Fig. 17 . Fig. 17. Open in a new tab Similar simulated granular materials. Optimization test of coal caving step distance Experimental method The 11508 fully mechanized caving face at Xiejiahegou Coal Mine operates with a shearer cutting depth of 0.8 m. To ensure a consistent total advance distance across different simulated conditions, three groups of caving step lengths and cycle numbers were selected for the experiments: (1) 12 × 2.7 cm; (2) 6 steps × 5.3 cm; and (3) 4 steps × 8.0 cm. A certain distance was set aside to prevent the impact of the boundary effect on both ends of the model. The main experimental procedure is as follows: The top coal drawing ports of supports No. 1 through No. 6 were opened sequentially, and a "gangue-sensing and door-locking" criterion was applied to each support. Weigh the top coal and gangue discharged from each support by electronic balance and electronic platform scale. Coal caving end, move the frame and repeat (1)–(2). Repeat (1)–(3) until the number of propulsions is reached. Migration of coal rock The initial caving in the top coal caving experiment under various caving step lengths results in a similar initial deformation of coal seam. Although the coal-rock interface morphology after initial caving exhibits some variation, the overall caving morphology is comparable. This is because the physical similarity simulation uses manual caving and the particle sizes are randomly distributed. Figure 18 illustrates the coal-rock interface morphology following the initial caving cycle. Fig. 18. Open in a new tab Experimental platform and instrument. Figure 18 shows that after the initial support movement and caving, the coal-rock interface exhibits an inverted arch form. Compared to larger step lengths (e.g., 1:2 mining-to-caving ratio), the smaller caving step length is more suitable for the initial top coal drawing because the immediate roof gangue falls directly behind the support tail beam after the initial support movement. Consequently, less top coal is left in the goaf, as the majority of the top coal is effectively recovered during the initial caving stage. When using large caving step lengths (e.g., the 1:3 mining-to-caving ratio), the top coal drawing space increases. The gangue behind it jumps into the top coal before it reaches the coal caving aperture. This premature entry of gangue leads to the remaining top coal being trapped above the flow path and unrecovered when the port is closed. Currently, the greatest amount of coal loss occurs during the first coal caving cycle and advancement step. Experimental data and analysis Prior to the experiment, the mass of the top coal and immediate roof particles for model construction were accurately weighed and quantified. Through theoretical calculation, the top coal mass contained within each caving step length is 1.69 kg, 3.32 kg, and 5.00 kg, respectively. Subsequently, the mass of discharged top coal and immediate roof particles was weighed and recorded at each step of the experiment. Tables 8 , 9 and 10 present the results, detailing the recovered mass of top coal and the mass of discharged gangue. Table 8. The total data table of one mining and one caving. Bracket number 1# 2# 3# 4# 5# 6# Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Excessive top coal Release amount Excessive gangue Release amount (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (%) (%) grand total 3236 1784 2801 1278 3468 1667 3635 1634 2790 1429 3161 1857 94.28% 33.57% Open in a new tab Table 9. Two mining and one caving total data table. Bracket number 1# 2# 3# 4# 5# 6# Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Excessive top coal Release amount Excessive gangue Release amount (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (%) (%) grand total 2766 1121 2399 1079 2987 979 2877 1319 2936 1153 1834 996 78.02% 29.61% Open in a new tab Table 10. Three mining and one caving total data table. Bracket number 1# 2# 3# 4# 5# 6# Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Coal release amount Discharge amount of gangue Excessive top coal Release amount Excessive gangue Release amount (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (%) (%) grand total 2295 1309 2130 836 2132 806 2603 746 2381 632 2526 619 69.47% 26.02% Open in a new tab From the data in Figs. 19 , 20 and 21 , it can be seen that the initial recovered mass of top coal is substantial across all caving step lengths, with the highest value reaching 436.74% of the theoretical mass. Furthermore, the initial recovered mass decreases as the caving step length increases. The total recovered mass of top coal for the 1st, 2nd, and 3rd caving step lengths are 19,091 g, 15,799 g, and 14,067 g, respectively. The corresponding recovery rates are 94.28%, 78.02%, and 69.47%; the gangue contents are 33.57%, 29.61%, and 26.02%. Fig. 19. Open in a new tab Coal and rock caving morphology after first moving support caving. Fig. 20. Open in a new tab Data diagram of each support of one mining and one caving. Fig. 21. Open in a new tab Data diagram of each support of two mining and one caving. During periodic caving, the mean recovered mass per caving cycle increases with the caving step length. The mean recovered mass for the 1st step length is 976.75 g, while the others are 1975.54 g and 2345.83 g, respectively. The calculated standard deviations for these cycles are 216.72 g, 879.82 g, and 192.00 g, respectively. Thus, the coal recovered in the 1st and 3rd step length conditions exhibits lower variance. When the caving space is small, the flow velocity of the top coal and gangue is relatively slow. If the gangue prematurely enters the flow path before the top coal is fully discharged, it leads to caving interruption. The gangue located above the support does not easily contaminate the flow path from the support tail beam to the port, allowing for more complete top coal recovery. However, the goaf-side gangue, being close to the drawing port, is prone to rapid ingress, causing caving termination. Consequently, the highest top coal recovery rate is achieved at the 1st step length, but this condition also yields the highest gangue content. When the caving space increases, the top coal above the tail beam collapses to the goaf floor, which isolates the goaf-side gangue. Under gravity, the flow velocity of the gangue above the support accelerates, causing it to intermix with the top coal and flow quickly to the drawing port. At this stage, the top coal recovery rate decreases, and the gangue content also decreases (as per the original text’s conclusion). Therefore, the 3rd step length yields the lowest top coal recovery rate but also the lowest gangue content. Combining these findings with the numerical simulation results, it is concluded that for the 11508 fully mechanized caving face at Xiejiahegou Coal Mine with a mining-to-caving ratio of 1:1.5, adopting the shortest caving step length (equivalent to "one mining and one caving") results in the highest top coal recovery rate and stable periodic discharge, despite a higher gangue content. Therefore, the optimal caving step length for the 11508 working face is determined to be the shortest length condition ("one mining and one caving"). Optimization test of coal caving sequence Experimental method The 11508 fully mechanized caving face at Xiejiahegou Coal Mine features a 1.6-m-wide hydraulic support. A 6 cm-wide movable stainless steel sheet was used in the experiment to replicate the support’s drawing aperture. The physical model was used to mimic top coal caving under three distinct caving procedures: multi-port sequential caving in a single round, single-port sequential caving in a single round, and single-port intermittent caving in a single round. The multi-port sequential caving test was not performed. This decision was made because the single-port sequential caving method is often used when the mining height is 1.96 m and the top coal density within the 7.1 m range of the operating face is lower than the density threshold for areas where the coal thickness exceeds 7.1 m. The primary procedure for the three simulated caving sequence experiments is as follows: Single-Port Sequential Caving in a Single Round:The top coal drawing is executed sequentially from support No. 1 to No. 6, and the drawing port is closed upon gangue emergence. Single-Port Intermittent Caving in a Single Round: The top coal drawing is executed sequentially in the order of supports No. 1, No. 3, No. 5, No. 2, No. 4, and No. 6. The drawing port is closed upon gangue emergence. Multi-Port Simultaneous Caving in a Single Round: The top coal drawing is executed sequentially by opening pairs of supports simultaneously: No. 1 and No. 2, then No. 3 and No. 4, and finally No. 5 and No. 6. The drawing ports are closed upon gangue emergence. Experimental data and analysis In constructing the physical model, the materials for the immediate roof and top coal were accurately weighed and quantified.The total mass of top coal above the simulated support area was determined to be 8.55 kg. During the experiment, the mass of top coal and gangue particles discharged at each caving cycle was weighed and recorded.The results, derived by calculating the mass of both the discharged top coal and gangue, are presented in Figs. 22 , 23 and 24 . Fig. 22. Open in a new tab Data diagram of each support of three mining and one caving. Fig. 23. Open in a new tab Single-round single-port sequential coal caving data. Fig. 24. Open in a new tab Single round single mouth interval coal drawing. As evident from the graphic, the total recovered mass of top coal for the three caving procedures—Single-Port Sequential, Single-Port Intermittent, and Multi-Port Sequential (all in a single round)—are 7,370 g, 5,270 g, and 6,810 g, respectively.The standard deviation of the recovered mass per support for these procedures are 214.12 g, 268.48 g, and 707.29 g, respectively. The corresponding recovery rates are 86.24%, 61.66%, and 79.68%; the gangue contents are 5.73%, 24.39%, and 24.83%. It is evident that the Multi-Port Sequential Caving procedure has a higher recovery rate (79.68%) than the Single-Port Intermittent Caving procedure (61.66%), with only a marginal difference in gangue content (24.83% vs. 24.39%). Therefore, the overall effect of Multi-Port Sequential Caving is marginally superior to that of Single-Port Intermittent Caving. However, the Single-Port Sequential Caving procedure exhibits the lowest gangue content (5.73%) and the most uniform recovered mass distribution per support (indicated by the smallest standard deviation, 214.12 g. Consequently, considering the 11508 working face operates with a mining-to-caving ratio of 1:1.5, the most appropriate primary caving procedure is Single-Port Sequential Caving, followed by Multi-Port Sequential Caving as the optimal secondary option. Similar simulation experiment of door closing principle The gangue-closure technique is widely employed during top coal caving operations in actual production. However, due to natural geological conditions and the caving process itself, gangue inevitably enters the drawing port prematurely, preceding the flow of top coal. This leads to either insufficient top coal recovery or excessive top coal losses. Specifically, while excessive coal caving releases higher-quality coal, it simultaneously increases the proportion of discharged gangue. Influenced by factors such as coal quality and subsequent coal preparation cost, we cannot solely focus on maximizing top coal recovery while neglecting the overall economic benefits. Therefore, after determining the reasonable coal drawing process parameters, it remains crucial to examine the internal relationship between top coal recovery rate and gangue content, refine the closure criterion, and balance coal resource recovery with economic benefits. Experimental method The principle of ceasing drawing upon gangue appearance is widely adopted in practical production. However, due to the influence of natural geological conditions and the specific drawing technology employed, gangue inevitably precedes the top coal and enters the draw opening, resulting in insufficient extraction of top coal and, consequently, loss of top coal resources. Although increasing drawing duration releases more top coal, it also significantly increases the amount of gangue extracted. Influenced by factors such as coal quality and subsequent washing costs, the focus cannot solely be on maximizing the top coal recovery rate while ignoring overall economic benefits. Therefore, even after determining reasonable drawing process parameters, it remains necessary to investigate the inherent relationship between the top coal recovery rate and the gangue content rate during the caving process. This research aims to define a clear cessation principle (draw-stopping criteria) that simultaneously considers both coal resource recovery and economic profitability. Using the bulk similar simulation platform, the experiment simulates the process where, after the gangue becomes visible, the top coal drawing continues. The initial moving assistance particles and the fragmented top coal pieces are then weighed and recorded at regular intervals. The main process of simulation experiment is as follows: Initial Drawing Phase: Sequentially perform the initial coal caving from Support 1# through 6#. Cease drawing at each support upon the observation of gangue, concluding with the operation of Support 6#. Phase I Continuous Drawing (5 s): Continue the drawing process in the 1# to 6# sequence. Set the drawing duration for each support to 5 s (5 s), continuing until the operation concludes at Support 6#. Phase II Continuous Drawing (2 s): Repeat the drawing process in the 1# to 6# sequence. Set the drawing duration for each support to 2 s (2 s), continuing until the operation concludes at Support 6#. Repeat step 3 ) 4 times. Perform the first support advance, and then repeat Steps 1 through 4. Experimental data and analysis During the model construction process, the immediate roof and coal seam materials were initially weighed and quantified. Based on calculations, the total mass of top coal overlying a single support was determined to be 10.79 kg during the initial drawing phase. Furthermore, after the first support advance, the total available top coal mass above the support was 3.85 kg (refer to Table 11 for detailed information). Table 11. Statistical data of excessive coal gangue. Coal caving stage Initial coal caving Caving coal after initial support moving Coal release amount Discharge amount of gangue Excessive top coal release amount Excessive gangue release amount Coal release amount Discharge amount of gangue Excessive top coal release amount Excessive gangue release amount Closing time (g) (g) (g) (g) (g) (g) (g) (g) See gangue 6340 191 1152 247 5 s 1660 1170 640 520 7 s 1410 1070 520 830 9 s 730 940 470 780 11 s 450 710 380 690 13 s 520 510 470 1000 Open in a new tab In the course of the experiment, the mass of immediate roof and top coal particles extracted at different stages—specifically, during the initial drawing, the continuous drawing phase, and the drawing following the first support advance—were carefully weighed and recorded. The collected data are summarized in Table 12 . Table 12. Statistical table of carbon and the rate at which excessive carbon caving occurs in the gangue outflow. Coal caving stage Initial coal caving Caving coal after initial support moving (%) (%) (%) (%) Closing time Coal caving rate Debris content Coal caving rate Debris content See gangue 58.74% 2.92% 39.52% 13.98% 5 s 37.28% 41.34% 27.69% 44.67% 7 s 50.48% 43.15% 30.92% 61.48% 9 s 52.79% 56.29% 40.45% 62.40% 11 s 68.92% 61.21% 54.92% 64.49% 13 s 256.22% 53.98% 150.67% 68.03% Open in a new tab In the initial coal drawing stage, a total of 11.11 kg of top coal and 4.69 kg of gangue were extracted. At the cessation point upon gangue appearance (the draw-stopping point), the maximum coal recovery rate reached 58.74%, with a gangue content rate of 2.92%. During the continuous drawing phase (over-drawing), the separation distance between the released top coal and gangue gradually decreased, indicating increasing mixing. This phenomenon is critical because the top coal drawing rate is generally positively correlated with the total amount of loose top coal available above the support throughout the caving time. After the first support advance, a total of 4.00 kg of top coal and 4.01 kg of gangue were extracted. At the cessation point, the top coal recovery rate was 16.74%, and the gangue content rate was 27.72%. The fluctuation in the quantity of top coal released per period was small. However, due to the higher overall top coal discharge in the initial drawing stage, the subsequent quantity of gangue released during the continuous drawing phase was larger, and the overall gangue output showed an increasing trend. Initial Drawing Stage: When the drawing duration reached 11 s (11 s) after gangue appeared, 10.60 kg of top coal was discharged, with 202.95 g remaining above the support. An additional 520 g of top coal, primarily near the support’s tail beam, was discharged between 11 and 13 s. Therefore, to ensure complete recovery of the accessible top coal above the support in this stage, it is sufficient to continue drawing for a total time of 13 s after gangue sighting. Drawing After Support Advance: During this phase, when drawing reached 11 s, 3.53 kg of top coal had been extracted, and 311 g remained above the supporting structure. An additional 470 g of top coal was released between 11 and 13 s, with approximately 16 g coming from the area in front of the support’s tail beam. Therefore, it is likewise sufficient to continue drawing for a total time of 13 s to fully recover the remaining top coal above the support. The determination of the closing principle The mass of top coal extracted up to a certain moment is defined as M, the mass of gangue extracted is G, the gangue content rate is H, and the ash percentage of the released material is S. Let the bulk density of coal be ρm = 1.44 g/cm 3 and the bulk density of gangue beρg = 2.5 g/cm 3 . The relationship between the ash content of the coal-rock mixture, the mass of extracted gangue, and the top coal mass can be obtained using the following formula: 14 15 In the initial drawing phase, if the complete recovery of all top coal above the support is set as the draw-stopping condition, drawing should continue for an additional 13 s after gangue sighting. At this point, the gangue content rate is H = 18.7% and the ash percentage is S = 33.2%. This corresponds to the discharged gangue volume accounting for approximately 18.7% (1/5) of the total volume, thus defining the “1/5 Gangue Cessation Principle.”Alternatively, if the criterion is based on the economical threshold (where the mass of extracted coal outweighs the mass of extracted gangue), drawing should continue for 7 s after gangue sighting. The resulting parameters are H = 12% and S = 26.72%. This means the gangue volume accounts for approximately 1/8 of the total volume, which is known as the “1/8 Gangue Cessation Principle.” In the cyclic drawing phase (or subsequent caving cycles), if complete top coal recovery is the criterion, drawing must continue for 13 s after gangue sighting. The parameters are H = 36% and S = 47.34%. This higher gangue content, approximately 36% (1/3) of the total volume, establishes the “1/3 Gangue Cessation Principle.”If the criterion is set at the economical threshold, drawing should continue for only 5 s after gangue sighting. The parameters achieved are H = 16% and S = 30.67%. The resulting gangue volume accounts for approximately 1/6 of the total volume, defining the “1/6 Gangue Cessation Principle.” Field validation of the 11508 longwall face However, the actual thickness of the coal seam in the 11508 longwall face is not constant. A caving ratio of 1:1.5 is only applicable when the working face cannot extract the full height in a single pass. The isoline map for the thickness of the No. 15 coal seam is presented in Fig. 25 . Fig. 25. Open in a new tab Single-motor serial coal extraction with numerous holes. As shown in the Fig, the thickness of the No. 15 coal seam is predominantly greater than 3.5 m. Furthermore, the top coal thickness increases significantly as it approaches the stop line. In this predominant area, a caving ratio of 1:1.5 can be employed for the retreat mining operation.However, there is also a small localized area where the top coal thickness ranges from 1.3 m to 3.5 m. Depending on the specific conditions in this area, the decision can be made to either adopt a full-height extraction in a single pass or maintain the 1:1.5 caving ratio. Conversely, when the top coal thickness is between 0.7 m and 1.3 m, only the full-height extraction in a single pass is permissible, and top coal caving should not be performed.Given that the area where the top coal thickness exceeds 3.5 m accounts for the majority of the working face, it is concluded that for the 11508 longwall face, the optimal top coal recovery effect is achieved when the caving step distance is 0.8 m and a caving ratio of 1:1.5 is adopted, corresponding to a cutting height of 1.96 m and a caved height of 2.94 m (Fig. 26 ). Fig. 26. Open in a new tab Isoline Map of the No. 15 Coal Seam Thickness. Table 13 presents the top-coal drawing data for an individual hydraulic support at the No. 11508 working face of Xiejiahegou Coal Mine. The data correspond to an operational scenario characterized by a cutting-caving ratio of 1:1.5 and a caving step distance of 0.8 m. Specifically, the cutting height is 1.96 m, and the caving height is 2.94 m. The caving process adopts the "one-cut-one-cave" method with an interval of 0.8 m. The width of a single hydraulic support is set at 1.5 m, in accordance with standard Longwall Top Coal Caving (LTCC) specifications. Table 13. Top coal drawing data for the No. 11508 working face. Cut coal quantity (t) Actual drawn top coal (t) Theoretical top coal reserves (t) Total output (t) Recovery rate (%) 3.39 4.36 5.08 7.75 91.5 Open in a new tab During field operations, the fragmentation and flow characteristics of the top coal remained highly stable. Drastic fluctuations, such as intermittent drawing failures followed by sudden mass surges between supports, were not observed. This stability is significantly advantageous for preventing blockages behind the hydraulic supports, controlling gangue content, and maintaining a consistent production rhythm. The steady coal flow, characterized by sufficient top coal fragmentation and minimal volumetric fluctuation, facilitates high-yield and high-efficiency mining. During the drawing process, the top coal caved uniformly with moderate lump sizes. No jamming caused by oversized blocks was observed, effectively ensuring the continuous patency of the coal drawing opening. Field observations indicate a good match between the drawing duration and the descending velocity of the top coal. Furthermore, high synergy between adjacent supports further enhanced the overall recovery efficiency. Consequently, the actual recovery rate reached 91.5%, approaching the theoretical limit. This validates the rationality and superiority of the parameter combination consisting of a 1:1.5 cutting-caving ratio and a 0.8 m drawing interval. Top coal drawing was conducted based on the caving step distances established in the aforementioned physical similarity simulation. For the No. 11508 working face of Xiejiahegou Coal Mine (with a defined width of 12.6 m), the caving step distances were optimized as follows: 0.81 m for the one-cut-one-cave method, 1.59 m for the two-cuts-one-cave method, and 2.4 m for the three-cuts-one-cave method. Table 14 presents the corresponding data for top coal drawing. Table 14. Top coal caving data diagram of the 11508 working face. Caving step distance (m) Cumulative discharged top coal quantity (t) Cumulative discharged gangue quantity (t) Top coal recovery rate (%) Gangue content rate (%) One mining one caving 0.76 343.6 277.9 94.28 44.7 Two minings one caving 1.59 284.4 191.4 70.28 40.2 Three minings one caving 2.32 253.2 142.5 69.47 36.0 Open in a new tab In the simulation of the 'single-cut caving’ method, the total coal output for the simulated section reached 343.6 tonnes, achieving a recovery rate of 94.28%. However, under the strict application of the 'close-on-gangue’ principle, the in situ gangue content reached 44.7%. This resulted in poor coal quality and placed significant pressure on subsequent washing and preparation processes. Conversely, when employing the 'three-cut caving’ method, the total output decreased to 253.2 tonnes—a loss of nearly 90 tonnes of coal—though the gangue content was reduced to 36.0%, yielding relatively better coal quality. As the caving interval increases, the volume of coal drawn per cycle rises, indicating that a larger interval provides greater expansion space, which facilitates top coal flow. However, excessive intervals lead to the premature caving of top coal above the hydraulic supports. Consequently, gangue tends to intrude into the discharge port before the top coal is fully drawn. This necessitates early closure of the gate to control quality, resulting in substantial loss of top coal. Considering the dual requirements for resource recovery and coal quality in field production, as well as the impact of washing costs on economic viability, an optimal balance between the caving interval and gangue content is required. Drawing upon research findings regarding the 'close-on-gangue’ principle from physical simulations, the protocol was refined for application at the 11508 working face, specifically targeting the high recovery characteristics of single-cut caving. During the initial caving stage, a '1/5 gangue appearance’ threshold was adopted for gate closure; this maintained a high drawing rate while keeping gangue content within an acceptable range. In the cyclic caving stage, a '1/6 gangue appearance’ threshold was flexibly applied based on caving behaviour and gangue intrusion trends, effectively mitigating the issue of excessive gangue inclusion caused by the over-pursuit of output volume. Field practice demonstrates that under this optimised 'close-on-gangue’ principle, the single-cut caving process not only maintained a top coal recovery rate exceeding 90% but also reduced gangue content from 44.7% to approximately 30%. This improved coal quality and reduced the pressure and costs associated with downstream processing. Simultaneously, the zoning of different cutting-caving ratios and the control of actual cutting and caving heights were effectively implemented. For areas with top coal thickness greater than 3.5 m, a cutting-caving ratio of 1:1.5 was strictly enforced (cutting height 1.96 m, caving height 2.94 m). For transition zones with top coal thickness between 1.3 m and 3.5 m, either single-pass full-height cutting or a 1:1.5 ratio was selected based on geological conditions and equipment capacity to ensure resource recovery and production continuity. In areas where top coal thickness was less than 1.3 m, single-pass full-height cutting was mandatorily adopted to avoid the resource waste and safety risks associated with top coal caving. The comprehensive application of these measures enhanced the overall mining efficiency and economic benefits of the 11508 working face, providing valuable practical experience for fully mechanized top coal caving in thick seams under similar conditions. Conclusion The study utilizes the 11508 working face of Xiejiahegou Coal Mine as its research context, focusing on the problem of low top coal recovery and high gangue content in shallowly buried, stable, thick LTCC (Longwall Top Coal Caving). The research methodology involves an integrated approach of field monitoring, theoretical analysis, numerical simulation, and physical modeling. The investigation focuses on optimizing technological parameters by analyzing the dynamic evolution of the coal-rock interface and the morphology of the top coal drawing body under various conditions (e.g., thickness, mining-to-drawing ratio, drawing step, and sequence). The objective is to explicitly define the draw-stopping criteria, providing a theoretical foundation for coordinating mining and drawing activities and selecting optimal parameters. The main conclusions are as follows: Combined with numerical simulation results, it is evident that top coal recovery is significantly influenced by the drawing sequence (caving scheme) and the mining-to-drawing ratio (mining step length). Impact of Drawing Sequence: The coal recovery rate, ranked from highest to lowest, follows the sequence: Two Mining/Two Drawing (2 M/2D) > One Mining/One Drawing (1 M/1D) > Two Mining/One Drawing (2 M/1D). Impact of Mining-to-Drawing Ratio: The maximum coal recovery rate, ranked from highest to lowest, is: 1:1.5 > 1:2 > 1:1 > 1:3. Based on the standard deviation of top coal mass across the sequential caving cycles, the quantity of top coal released between each phase shows the highest uniformity (is best balanced) when the ratio is 1:1.5. Conversely, the discharge amount exhibits the most noticeable fluctuation when the mining-to-drawing ratio is 1:3. The discharge uniformity for the 1:1 ratio is also comparatively well-distributed. Beyond the optimization of the caving step distance and sequence, the top coal drawing step distance is closely related to the size of the drawing area. When the drawing space is small (short step distance), gangue has difficulty entering the drawing opening from the support’s tail beam due to the slow movement of the coal and gangue. However, even if the top coal is fully drawn, gangue from the adjacent goaf side can easily inrush and obstruct the drawing opening.. When employing the 1 M/1D (one mining/one drawing) scheme, the gangue from the goaf side is partially isolated, potentially resulting in the highest top coal recovery rate and the lowest gangue content rate. As the drawing area expands, the top coal above the tail beam falls to the goaf floor. This accelerates the flow of gangue located behind the support, causing it to mix with the top coal and flow into the drawing opening. Consequently, both the gangue content rate and the top coal recovery rate decrease. However, the top coal recovery rate is the lowest, while the gangue content rate remains low. As the gangue content difference is not statistically significant but the coal drawing rate is higher, sequential multi-port drawing (e.g., 2 M/2D) shows a slightly superior performance compared to single-port cyclic drawing. At the optimal drawing rate, the gangue content is lowest, and the quantity of coal drawn between support is best matched. Therefore, for the coal seam at the 11508 working face, when the mining-to-drawing ratio is 1:1.5, the 1 M/1D scheme is recommended. The most reasonable overall strategy is to first utilize single-port sequential drawing, followed by cyclic multi-port drawing. The results from the physical modeling experiment reveal that the appropriate draw-stopping criteria depend on the drawing phase and objective. If complete recovery of the top coal above the support is the primary cessation criterion, the “1/5 Gangue Cessation Principle” is applied. If the criterion is set at the economical threshold (i.e., the mass of liberated coal exceeds the mass of gangue), the “1/8 Gangue Cessation Principle” is utilized. If the complete recovery of the available top coal above the support is the criterion, the “1/3 Gangue Cessation Principle” is employed. If the criterion is based on the economical threshold, the “1/6 Gangue Cessation Principle” is applied to optimize top coal recovery. Author contributions Shixiong Wu:Conceptualization, Methodology, Software; Xun Xu: Data curation, Writing- Original draft preparation; Jun Wang: Visualization, Investigation; Dezhong Kong: Visualization, Investigation; Guiyi Wu:Supervision; Qinzhi Liu: Software, Validation; Yunjun Zuo:Project administration. Funding Qiankehe Platform Talents, Grant/Award Number: (GCC [2023] 056); Guizhou Provincial Department of Education 2023 Annual College Science and Technology Innovation Team(Guizhou Education Technology [2023] 055); National Natural Science Foundation of China (No.52564007 , No. 52164002, No. 52164005 ); Guizhou Provincial Basic Research Program(Natural Science) (Qianke He Foundation-ZK[2024] Key 022); Guizhou Provincial Science and Technology Department Innovation Talent Team Construction Project (Qiankehe talent CXTD [2025]025);National Natural Science Foundation of China (52404117), Guizhou Provincial Science and Technology Projects ([2024] Youth 138, KXJZ[2024] 020), Guizhou Provincial Basic Research Program (MS [2025] 631), Guizhou University talent project ([2023] 35), Guizhou University Basic Research Program ([2024] 19). Data availability All data, models, and code generated or used during the study appear in the published article. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Jun Wang, Email: [email protected]. Dezhong Kong, Email: [email protected]. References 1. Jiachen, W. Engineering practice and theoretical progress of top coal caving mining in China. Coal J. 43 (01), 43–51 (2018). [ Google Scholar ] 2. Liu, J., Chen, S. L., Wang, H. J., Li, Y. C. & Geng, X. W. The migration law of overlay rock and coal in deeply inclined coal seam with fully mechanized top coal caving. J. Environ. Biol. 36 (4), 821–827 (2015). [ PubMed ] [ Google Scholar ] 3. Chen, L. et al. Analysis of the impact of key strata failure on overlying strata and surface transport patterns. Eng. Fail. Anal. 10.1016/j.engfailanal.2024.108532 (2024). [ Google Scholar ] 4. Klishin, V. I., Fryanov, V. N., Pavlova, L. D. & Opruk, G. Y. Modeling top coal disintegration in thick seams in longwall top coal caving. J. Min. Sci. 55 (2), 247–256. 10.1134/S1062739119025527 (2019). [ Google Scholar ] 5. Zhang, Y., Xu, M., Kong, D., Wu, G. & Song, G. Study on the fracture-orientation effect of hard roof breakage characteristics in Karst Mine Mining. Int. J. Numer. Anal. Methods Geomech. 10.1002/nag.70138 (2025). [ Google Scholar ] 6. Wen, H. et al. Study on overlying strata breaking law and dynamic pressure characteristics of shallow buried coal seam mining under gully landform. Eng. Fail. Anal. 183 , 110276. 10.1016/j.engfailanal.2025.110276 (2026). [ Google Scholar ] 7. Wang, B. et al. Failure mechanism of limestone floor in coal seams under water-rock interactions: A Case study from northern Guizhou, China. Eng. Fail. Anal. 184 , 110350. 10.1016/j.engfailanal.2025.110350 (2026). [ Google Scholar ] 8. Zhang, Y. et al. Study on the evolution and transport pattern of overburden fracture induced by mining in cave mines. Sci. Rep. 10.1038/s41598-025-17438-8 (2025). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Li, Q., Yang, S. L., Yue, H. & Kong, D. Z. Study on caving characteristics of roof and floor and law of ground pressure behavior in near-vertical coal seams mining: a three-dimensional similarity simulation experiment. Rock Mech. Rock Eng. 10.1007/s00603-025-04922-w (2025). [ Google Scholar ] 10. Pan, X. et al. Development, optimisation and performance prediction of a novel cement-based materials for borehole sealing. Construct. Build. Mater. 10.1016/j.conbuildmat.2025.141404 (2025). [ Google Scholar ] 11. Kong, D. Z., Jiang, W., Chen, Y., Song, Z. Y. & Ma, Z. Q. Study of roof stability of the end of working face in upward longwall top coal. Arab. J. Geosci. 10.1007/s12517-017-2978-0 (2017). [ Google Scholar ] 12. Song, Z. Y. & Konietzky, H. A particle-based numerical investigation on longwall top coal caving mining. Arab. J. Geosci. 10.1007/s12517-019-4743-z (2019). [ Google Scholar ] 13. Zhang, J. W., Wang, J. C., Wei, W. J., Chen, Y. & Song, Z. Y. Experimental and numerical investigation on coal drawing from thick steep seam with longwall top coal caving mining. Arab. J. Geosci. 10.1007/s12517-018-3421-x (2018). [ Google Scholar ] 14. Liu, Y. H. et al. The migration and evolution law of overlying strata and the instability and failure characteristics of end face roof under the condition of ascending mining in close distance coal seam: Case study. Eng. Fail. Anal. 10.1016/j.engfailanal.2024.108809 (2024). [ Google Scholar ] 15. Yang, L., Wang, J. C., Yang, S. L., Li, L. H. & Wu, S. X. Improving the top coal recovery ratio in longwall top coal caving mining using drawing balance analysis. Int. J. Min. Reclam. Environ. 37 (5), 319–337. 10.1080/17480930.2023.2184036 (2023). [ Google Scholar ] 16. Yu, B., Zhao, J. & Xiao, H. T. Case study on overburden fracturing during longwall top coal caving using microseismic monitoring. Rock Mech. Rock Eng. 50 (2), 507–511. 10.1007/s00603-016-1096-8 (2017). [ Google Scholar ] 17. Wang, K., Zhao, T., Yetilmezsoy, K. & Zhang, X. Q. Cutting-caving ratio optimization of fully mechanized caving mining with large mining height of extremely thick coal seam. Adv. Civil Eng. 10.1155/2019/7246841 (2019). [ Google Scholar ] 18. Zhang, Y. et al. Influence of refuse content on economic benefits for fully mechanized top coal caving. Procedia Eng. 26 , 2391–2399. 10.1016/j.proeng.2011.11.2450 (2011). [ Google Scholar ] 19. Cao, J. et al. A study of the top-coal-drawing law of steeply inclined and extremely thick coal seams in the Wudong Coal Mine. Processes 10.3390/pr12030481 (2024). [ Google Scholar ] 20. Liang, Y. et al. Study on roof-coal caving characteristics with complicated structure by fully mechanized caving mining. Shock Vib. 10.1155/2019/6519213 (2019). [ Google Scholar ] 21. Jiang, H. Y., Song, Q., Gao, K., Song, Q. J. & Zhao, X. G. Rule-based expert system to assess caving output ratio in top coal caving. PLoS ONE 10.1371/journal.pone.0238138 (2020). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Wang, J., Yang, S., Li, Y., Wei, L. & Liu, H. Caving mechanisms of loose top-coal in longwall top-coal caving mining method. Int. J. Rock Mech. Min. Sci. 71 , 160–170. 10.1016/j.ijrmms.2014.04.024 (2014). [ Google Scholar ] 23. Liu, C. & Li, H. Numerical simulation of realistic top coal caving intervals under different top coal thicknesses in longwall top coal caving working face. Sci. Rep. 10.1038/s41598-021-92624-y[ (2021). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Zhao, T. L. et al. CDEM simulation of top-coal caving characteristics under synchronous multiwindow drawing in longwall top-coal caving mining of extra-thick coal seam. Energy Sci. Eng. 10 (9), 3617–3637. 10.1002/ese3.1245 (2022). [ Google Scholar ] 25. Zhang, Q. L., Yue, J. C., Liu, C., Feng, C., & Li, H. M. Study of automated top-coal caving in extra-thick coal seams using the continuum-discontinuum element method. Int. J. Rock Mech. Min. Sci. 122 (2019). 26. Le, T. D., Mitra, R., Oh, J. & Hebblewhite, B. A review of cavability evaluation in longwall top coal caving. Int. J. Min. Sci. Technol. 27 (6), 907–915 (2017). [ Google Scholar ] 27. Le, T. D., Oh, J., Hebblewhite, B., Zhang, C. G. & Mitra, R. A discontinuum modelling approach for investigation of Longwall Top Coal Caving mechanisms. Int. J. Rock Mech. Min. Sci. 106 , 84–95 (2018). [ Google Scholar ] 28. Yang, L., Li, L. H., & Wei, W. J. Optimization of caving technology in an extrathick seam with longwall top coal caving mining. Adv. Mater. Sci. Eng. (2021). 29. Zhang, Y., Liu, J. K., Li, Q., Zhao, J. J. & Ma, Y. Study on the process optimization of fully-mechanized hard top-coal caving with large inclination angle. AMR 889–890 , 234–240. 10.4028/www.scientific.net/amr.889-890.234 (2014). [ Google Scholar ] 30. Song, Z. Y., Konietzky, H., & Herbst, M. Drawing mechanism of fractured top coal in longwall top coal caving. Int. J. Rock Mech. Min. Sci. 130 (2020). 31. Zhang, N. B., Liu, C. Y. & Yang, P. J. Flow of top coal and roof rock and loss of top coal in fully mechanized top coal caving mining of extra thick coal seams. Arab. J. Geosci. 9 , 6 (2016). [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Data Availability Statement All data, models, and code generated or used during the study appear in the published article. Articles from Scientific Reports are provided here courtesy of Nature Publishing Group ACTIONS View on publisher site PDF (8.0 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top