Study on Coal Fragmentation Induced by Instantaneously Depressurized Gas and Its Influence on Coal and Gas Outburst: A Case Study of Different Gas Types

Abstract

Coal and gas outburst, as an extremely destructive underground disaster, poses serious threats to mine production safety and global energy supply. The mechanisms of this disaster, particularly how gas participates in and affects coal mass fragmentation, have not been fully revealed. To investigate this issue, this study simulated the coal-breaking process through instantaneously releasing high-pressure gas saturated in coal samples under gas–stress coupled conditions, employed image binarization method to quantitatively analyze the deformation and fragmentation characteristics of coal samples under different gas conditions, and conducted corroborative analysis from mechanical response and expansion energy perspectives. The results demonstrated that with the enhancement of gas adsorptive ability, gas desorption rate and amount accelerated, carried energy increased, and the long-term adsorption-induced degradation became more significant, resulting in greater extents of coal sample damage. Additionally, a rarely reported axial stress rebound phenomenon was observed, where axial stress underwent rapid decline followed by swift recovery to nearly initial levels within extremely short timeframes. This indicated that the instantaneously depressurized gas-induced coal fragmentation in coal seams level intensifies with the enhancement of adsorptive ability of different gases. The findings of this study may be helpful for understanding the gas participating in coal–rock damage during outburst occurrences, further ensuring mine safety production and global energy security.

1. Introduction

With urban development, global energy demand continues to grow, and energy requirements of various countries are constantly increasing [1,2,3]. Coal, as a solid combustible mineral, has been one of the primary energy sources since the eighteenth century [4,5]. To this day, the demand for coal in various countries remains high [6,7,8]. As coal mining depth and intensity continue to increase, mine gas conditions become increasingly complex. Consequently, the probability and frequency of coal and gas outburst disasters have increased dramatically [9,10]. China is a major coal-producing country and the one with the most frequent accidents, with hundreds of fatalities from coal mine accidents occurring annually [11,12,13]. Among these, coal and gas outburst accidents are the main type of accidents causing casualties and economic losses to coal mine workers in China [14,15].

Coal and gas outburst, as a complex type of coal–rock dynamic disaster, surges large amounts of coal dust and generates high concentrations of gas in an extremely short time when outburst disasters occur [16,17,18]. Research on the occurrence mechanism of coal and gas outburst accidents has always been of paramount importance, which is significant for both mine safety production and production efficiency [19]. Numerous scholars have conducted extensive research on this topic [20,21,22,23,24,25,26]. The “comprehensive action hypothesis” proposes that outburst disasters are influenced by the combined effects of in situ stress, gas, and the physical and mechanical properties of coal. This theory can explain outburst mechanisms more objectively and comprehensively and has been widely accepted [27,28,29]. In the early stage, research by Beamish et al. [30] indicated that coal and gas outburst disasters resulted from the comprehensive interaction of factors including stress, gas, and the physico-mechanical properties of coal and rock, with coal–rock type playing a crucial role. Subsequently, Sobczyk [31] experimentally investigated the influence of adsorption processes on gas pressure using carbon dioxide and nitrogen, and the results showed that adsorption processes might promote the occurrence of outbursts. Building on this foundation, the authors conducted subsequent experiments [32] using laboratory simulation equipment for outburst disasters, comparing the differences in outburst initial conditions under different gas conditions, which further confirmed that outburst initial conditions were determined by multiple factors, collectively. In recent years, Geng et al. [33] achieved the first synchronous observation of the three-dimensional dynamic evolution patterns of gas pressure field, temperature field, and coal stress field during the outburst process, validating the synergistic mechanism of “elastic strain energy-gas expansion energy”. Following this, research by Lei et al. [34] revealed that outburst occurrence is not an instantaneous process but depends on fracture development, where increased porosity allows free gas energy to accumulate sufficiently to trigger outbursts, with gas desorption contributing most of the energy in coal fragmentation during outbursts. Recently, research by Soleimani et al. [35] proposed that the essence of outbursts was an energy imbalance process involving elastic strain energy, gas expansion energy, coal fragmentation energy, and kinetic energy.

Overall, numerous scholars have conducted in-depth research on coal and gas outbursts and proposed various theoretical hypotheses, providing important theoretical guidance for its prevention and control. However, these hypotheses were all proposed for specific conditions and still have certain limitations in providing a unified mechanism explanation. Nevertheless, the comprehensive action hypothesis and energy hypothesis have been widely recognized and applied in related research, and the viewpoint that coal instability and fragmentation serve as a necessary condition for outburst occurrence has been clearly established. Therefore, studies on coal fragmentation mechanisms under gas–stress coupled conditions during coal and gas outburst occurrence possess much theoretical significance and practical value.

Coal mass failure is the prerequisite for coal and gas outburst occurrence and continuous coal breaking to maintain outburst development. Focusing on this core scientific issue, numerous scholars have conducted extensive research on coal fragmentation under fluid–solid coupling and achieved fruitful results. Among them, Tu et al. [36] employed uniaxial compression tests to measure and comparatively analyze the fragmentation work of intact coal samples. Concurrently, research by Hu et al. [37] focused on “stress-disturbance-fragmentation” correlations and verified through true triaxial testing systems that high energy input led to significant increases in fragmentation degree. From an energy perspective, research by Shu et al. [38] demonstrated that free gas expansion energy constituted the main source of coal fragmentation initiation energy, while gas adsorption–desorption energy provided energy support for continuous coal dust ejection. Similarly, research by Shi et al. [39] confirmed that outburst coal fragmentation mainly originated from coal elastic potential energy and gas expansion potential energy, with gas expansion energy accounting for more than 70%. Meanwhile, Wang et al. [40] conducted simulation experiments under different gas pressure conditions through true triaxial tests and found that increased gas pressure within a certain range promoted crack propagation. Furthermore, Bei et al. [41] studied the influence patterns of confining pressure stress on coal–rock fragmentation, with results showing that high stress increased energy accumulation, leading to increased proportions of small particles after fragmentation and larger fractal dimensions. Zhou et al. [42] provided microscopic evidence for the chain reaction mechanism in coal and gas outbursts through instantaneous release tests of adsorbed pressured gas, where the instantaneous release triggered outburst and provided energy for coal fragmentation. Finally, research by Guo et al. [43] proved that gas collectively weakens coal strength through adsorption and desorption effects, making coal more prone to be fragmented. Based on these fundamental understandings of coal fragmentation mechanisms and gas–solid coupling processes, numerous scholars have conducted methane pre-drainage technology research to achieve effective prevention and control of coal mine disasters, providing important technical support for deep mining safety [44,45].

In summary, the occurrence of coal and gas outbursts is a gas–solid coupling physical process influenced by multiple factors. The fragmentation of coal during outbursts results from the combined effects of multiple factors, among which the presence of gas plays an important role. Domestic and international scholars have conducted extensive research on coal fragmentation and the mechanism of coal and gas outbursts. However, existing research focused mainly on the state characteristics before and after coal fragmentation and the distribution characteristics of fragmented coal particles, while research on the fragmentation mechanism of outburst-coal under gas–solid coupling conditions and its influence on outburst occurrence remains insufficient. This study investigated the instantaneously depressurized gas-induced coal fragmentation characteristics under different gas type conditions, which is of great significance for understanding the mechanism of coal and gas outburst and has far-reaching implications for underground coal mine safety production and global energy security.

2. Materials and Methods

2.1. Coal Sample Preparation

The fresh coal samples used in the experiments were all collected from the coal mining face of No. 5 coal seam in Panzhihua Coal Mine located in southwestern China. Raw coal samples exhibit mechanical behavior and brittle failure characteristics that are closer to actual field conditions. These characteristics are more consistent with the sudden characteristics of coal and gas outbursts, facilitating better simulation of in situ coal failure scenarios. Therefore, the selected outburst-prone raw coal samples in this study were from a bituminous coal seam belonging to the Upper Permian Longtan Formation (P₃l). The coal sample used in the experiment is shown in Figure 1.

First, proximate analysis was conducted on coal samples using a 5E-MACIII infrared rapid coal analyzer (Kaiyuan Instruments, Changsha, China). Approximately 10 g of coal samples (particle size < 80 mesh) were dried in a vacuum oven for 12 h at 80 °C. Then, 1 g of the dried coal sample was taken for testing, and the results are shown in

Table 1. Proximate analysis results for coal samples [46].

Coal Type	Visual Density (g/cm3)	Real Density (g/cm3)	Porosity (%)	Aad (%)	Vdaf (%)	Mad (%)	FCad (%)
Bituminous coal	1.405	1.485	9.16	14.93	17.28	1.10	66.73

A_ad represents the ash content on air-dried basis, V_daf represents the volatile matter content on dry-ash-free basis, M_ad represents the moisture content on air-dried basis, and FC_ad represents the fixed carbon on air-dried basis.

.

Then, the adsorptive ability of prepared coal samples to different gases was determined through isothermal adsorption experiment. The isothermal adsorption experimental data were fitted according to the Langmuir equation, and the Langmuir adsorption constants of the coal samples for three gases were obtained, as detailed in

Table 2. Langmuir adsorption constants of the coal samples for three gases.

Gas Types	Langmuir Adsorption Constant a (cm3/g)	Langmuir Adsorption Constant b (MPa−1)
N2	9.42	3.61
CH4	17.46	1.93
CO2	37.30	1.22

.

Additionally, coal is a porous medium rich in numerous pores and fissures and is a typical anisotropic material. The basic mechanical properties of coal samples were tested. To obtain more accurate and reliable uniaxial compressive strength values for the raw coal samples, a total of four groups of raw coal samples were used for uniaxial compression tests, and the obtained results are shown in

Table 3. Results of the uniaxial compression tests for raw coal samples [46].

Coal Sample	$\mathbf{Uniaxial}\mathbf{Compression}\mathbf{Strength}{\mathit{\sigma }}_{\mathit{c}}$ (MPa)	$\mathbf{Mean}\mathbf{Value}\mathbf{of}{\mathit{\sigma }}_{\mathit{c}}$ (MPa)	$\mathbf{Elastic}\mathbf{Modulus}\mathit{E}$ (GPa)	$\mathbf{Mean}\mathbf{Value}\mathbf{of}\mathit{E}$ (GPa)	$\mathbf{Poisson’s}\mathbf{Ratio}\mathit{\nu }$	$\mathbf{Mean}\mathbf{Value}\mathbf{of}\mathit{\nu }$
RC-1	6.22	5.63	0.867	1.014	—	0.310
RC-2	5.57		0.923		—
RC-3	5.89		1.255		0.32
RC-4	4.83		1.011		0.30

.

A total of five groups of conventional triaxial compression tests were conducted on raw coal samples, the cohesion ($c$) and the internal friction angle ($\varphi$) of raw coal samples calculated based on the experimental results are shown in

Table 4. Results of cohesion force and internal friction angle of raw coal samples under different confining pressures [46].

Coal Sample	$\mathbf{Confining} \mathbf{Pressure} {\mathit{\sigma }}_2={\mathit{\sigma }}_3$ (MPa)	$\mathbf{Cohesion}\mathbf{Force}\mathit{c}$ (MPa)	$\mathbf{Mean}\mathbf{Value}\mathbf{of}\mathit{c}$ (MPa)	$\mathbf{Internal}\mathbf{Friction}\mathbf{Angle}\mathit{\varphi }$ (°)	$\mathbf{Mean}\mathbf{Value}\mathbf{of}\mathit{\phi }$ (°)
RT-1	2	0.94	1.35	60.94	49.43
RT-2	4	1.27		51.62
RT-3	5	1.41		47.69
RT-4	6	1.59		42.84
RT-5	8	1.54		44.06

, with values of 1.35 MPa and 49.43°, respectively.

2.2. Experimental Procedure

2.2.1. Instantaneously Depressurized Gas-Induced Coal Fragmentation Tests (IDG-CF Tests)

The apparatus for the coal-breaking simulation of outbursts with symmetrical openings [47], as shown in Figure 2, from the State Key Laboratory of Coal Mine Disaster Dynamics and Control of Chongqing University, was employed to conduct IDG-CF tests under axial loading conditions. This apparatus mainly consists of a coal sample chamber, a loading unit, a gas release unit, and a real-time data acquisition unit. It can withstand a maximum axial load of 1000 kN, confining pressure of 60 MPa, and gas pressure of 6 MPa. The apparatus was designed to ensure uniform instantaneous gas release, thereby excluding the influence of unidirectional pressure impact on coal breaking. Thus, it can effectively simulate the instantaneous characteristics of outbursts and is applicable to samples of various sizes, providing a reliable experimental platform for studying the outburst-coal fragmentation mechanism, especially the influence of instantaneously depressurized gas. The scheme of IDG-CF tests is detailed in

Table 5. Experimental scheme of IDG-CF tests.

Coal Sample Type	Coal Sample Size			Gas Type	Bursting Disc Pressure (MPa)	Gas Pressure (MPa)	Axial Stress (MPa)
	Diameter (mm)	Height (mm)	Mass (g)
Raw coal samples	98.63	59.90	653.70	N2	4	3.5	5.7
	98.56	59.64	657.70	CH4	4	3.5	5.7
	99.02	59.06	656.30	CO2	4	3.5	5.7

.

The bursting disc, with a rated pressure of 4 MPa, was used to simulate the instantaneous characteristics. Three gases, nitrogen (N₂), methane (CH₄), and carbon dioxide (CO₂), were separately injected for adsorption with an equilibrium pressure of 3.5 MPa. An axial stress of 5.7 MPa, 78% of the average uniaxial compression strength, was applied. The pressurized gas was instantaneously released through the symmetrical four openings.

2.2.2. Procedures of IDG-CF Tests

① Measure and record dimensions and mass of testing coal samples (see Table 5). Both end faces (A and B faces) and the lateral surface (C) of coal samples were divided into four corresponding regions, sequentially labeled as regions 1, 2, 3, and 4, facilitating the observation of coal sample failure after testing, as shown in Figure 3.

② Take photographs of each region of coal samples before testing for comparative analysis after tests.

③ Attach strain gauges to coal surface: two gauges for axial deformation monitoring were along the axial direction on regions 1 and 4, and another two for lateral deformation were along the circumferential direction on regions 2 and 3.

④ Position coal samples at the sample chamber center, connect strain gauges to an ASMD3 resistance strain monitor (Beijing Asmik Technology, Beijing, China), and ensure strain data acquisition functionality is confirmed.

⑤ Install bursting disc and seal the entire gas release unit; specifically, the bursting disc was sealed using a 704 room-temperature-vulcanizing silicone rubber. Then, connect gas pressure and temperature sensors.

⑥ Vacuumize the entire test system to −10 kPa for at least 12 h to remove gases from the sample chamber and coal sample pores, eliminating the influence of non-test gases.

⑦ After the gas-tightness of the sample chamber is checked, inject test gas slowly into the sample chamber until the required pressure is achieved. Maintain the gas pressure for at least 168 h to allow coal samples to full adsorb gas to equilibrium state, as the monitored gas pressure remains unchanged (gas can be properly supplemented).

⑧ Connect stress sensor, apply the axial stress and maintain for 2 h to allow stress redistribution within coal samples. Subsequently, further inject test gas to exceed the bursting disc rated pressure, triggering disc failure to achieve instantaneous gas release.

⑨ Save and analyze test data, take photographs of after-test sample, and observe coal sample failure. The detailed loading paths of gas pressure and axial stress are shown in Figure 4.

2.3. Determination of Characteristic Crack Parameters Using Image Binarization Method

The image binarization method, as an effective image preprocessing technique, has extensive applications in coal sample surface crack identification [48]. To more clearly and quantitatively analyze and compare the damage extents of raw coal samples before and after tests, this study employed the image binarization method and made some customized development to extract surface crack information and determine crack characteristics, conducing to in-depth quantitative analysis and exploration into macroscopic manifestations of gas–stress coupled effects on coal fragmentation during outbursts. To facilitate the extraction of key information from images, color images are generally subjected to grayscale processing before binarization. This study used coal–rock crack visualization tools to perform binarization processing on coal sample photographs before and after IDG-CF tests. The specific method is shown in Figure 5.

Therefore, we can define the planar porosity of the coal samples as the ratio of total area of surface cracks extracted after binarization ($S$) and specimen cross-sectional area ($S_0$), as the following formula:

$$\phi ={\displaystyle \frac{S}{S_0}}={\displaystyle \frac{{\displaystyle \sum _i^m{a}_i^2}}{{\displaystyle \sum _i^M{S}_i}}}={\displaystyle \frac{{\displaystyle \sum _i^{m}n{\delta }^2}}{{\displaystyle \sum _i^M{\delta }^2}}}$$

(1)

where $S_i$ is the area of a single pixel; $m$ is the number of pores on the binarized specimen surface; $M$ is the number of pixels required to cover the entire specimen cross-section; and $a_i$ is the side length of the $i$-th pore.

$S$, $S_0$, and the area $S_i$ of each pore pixel block can be calculated using the built-in program of the coal–rock crack visualization tool, based on which the characteristic size of cracks can be calculated. As illustrated in Figure 6, make $d_{j,i}$ represent the characteristic size of the $j$-th crack, whose length is the diagonal length of the square pore pixel block. The characteristic size $d_{j,i}$ can be approximated as the crack aperture, with the value

$$d_{j,i}=\sqrt{2}{a}_i=\sqrt{2S_i}$$

(2)

If the total number of cracks on the coal sample surface is $N$, the cracks on the surface of the coal sample could be considered being formed by $k$ interconnected pore pixel blocks, then we can obtain the average characteristic size of surface cracks, which is

$$D_c={\displaystyle \frac{1}{N\cdot k}}{\displaystyle \sum _j^N{\displaystyle \sum _i^k{d}_{j,i}=}}{\displaystyle \frac{\sqrt{2}}{N\cdot k}}{\displaystyle \sum _j^N{\displaystyle \sum _i^k\sqrt{S_i}}}$$

(3)

where $i$ is the set of pores, and $j$ is the set of cracks.

Equations (2) and (3) can be customized and embedded into the coal–rock crack visualization tool to calculate the characteristic size and its distribution. Additionally, the crack length $L$ and the fractal dimension $D$ of surface cracks can be measured using the built-in tools of the software.

3. Results and Discussion

3.1. Coal Sample Damage Analysis After IDG-CF Tests

IDG-CF tests on raw coal samples under different gas conditions were conducted according to the experimental scheme, and the test results are shown in Figure 7. It demonstrated that coal samples suffered different degrees of damage under different gas conditions when high-pressure gas was instantaneously released after adsorption equilibrium was reached. As for the high-pressure N₂ conditions, observations from surfaces A and B revealed that the original cracks underwent significant propagation, but the samples basically maintained intact boundaries. Observations from surface C showed that the coal samples experienced obvious coal particle spalling and splitting, forming distinct new surfaces and micro-cracks.

In comparison, the damage to coal samples tested under high-pressure CH₄ conditions was more severe. The boundary of region A was damaged with a small amount of coal particle spalling. In region B, numerous original cracks underwent significant propagation, forming a large number of new micro-cracks. The boundaries were no longer intact, with clear new boundaries formed by coal particle spalling. Similarly, surface C also suffered severe damage with large-area coal particle spalling, about 40% new surfaces, and some sharp fractures being formed.

The damage to coal samples tested under high-pressure CO₂ conditions was most intense. In region A, the boundaries were incomplete, with a small amount of coal particle spalling and numerous new cracks, which were interconnected after large-area new surfaces forming and massive coal block spalling were formed in region C. Moreover, region C underwent splitting along bedding weak planes and formed new interconnected cracks, with damage occurring based on existing defects, further forming new sharp damage interfaces, which occupied about 30% of total area.

Overall, preliminary appearance observations indicated that the degree of coal sample damage was related to the adsorptive capacity of the test gases, with stronger adsorptive ability of the test gas resulting in more severe damage to the coal samples.

3.2. Analysis of the Cracks Distribution Characteristics on Coal Sample Surface After IDG-CF Tests

To further analyze the effect of different gases on outburst-coal fragmentation, crack information from coal sample photographs before and after tests was extracted, characteristic crack parameters were compared, and the damage extents of coal samples was quantitated. The crack information extraction results are shown in Figure 8. The portions shown in the figure were the local feature extractions from obviously damaged regions in Figure 7, where black represented coal matrix and colors represented the cracks with various characteristic sizes ($d_{j,i}$). The binarized results of coal sample cracks before and after tests under different gas conditions are summarized in

Table 6. Data statistic results of binarized coal sample cracks before and after tests under different gas conditions.

Gas Type	Cracks Before IDG-CF Tests	Cracks After IDG-CF Tests	Propagation Degree
N2	Original cracks with size of 15.2 mm, 18.93 mm	Propagated to 18.8 mm, 19.46 mm; Newly generated cracks with size of 14.56 mm; Small amount of regional newly generated cracks.	2.7–23.7%
CH4	Original cracks with size of 10.17 mm, 14.89 mm, 13.72 mm, 12.55 mm	Propagated to 33.39 mm, 21.73 mm, 15.77 mm, 27.61 mm; Numerous regional newly generated cracks.	14.9–228.3%
CO2	Small amount of local primary cracks	Newly generated cracks with size of 18.04 mm, 18.18 mm, 29.98 mm, 26.44 mm, 47.06 mm, 37.96 mm, 27.35 mm, 64.11 mm; Numerous new surfaces.	—

.

As shown in the figure, coal sample surfaces were distributed with original cracks of different sizes before testing under high-pressure N₂ conditions, where the lengths of visible micro-cracks in regions A-1 and B-1 were 15.2 mm and 18.93 mm, respectively. After tests, cracks in region A-1 underwent propagation, with lengths extending by approximately 23.7% to 18.8 mm. The original micro-cracks in region B-1 became connected, expanding by approximately 2.7% to 19.46 mm. In addition, regional cracks with lengths of about 15 mm were newly formed after coal particle spalling at the B-1 boundary. Moreover, numerous new cracks in region C-2 and macroscopic regional cracks with lengths of about 100 mm in region C-1 were also generated with coal particle spalling, numerous new crack formations, and original crack propagations.

Under high-pressure CH₄ conditions, original cracks on coal sample surfaces were relatively well-developed before tests. The lengths of visibly macroscopic original cracks in region B-3 were 10.17 mm, 14.89 mm, and 13.72 mm, respectively. They were significantly propagated with lengths extending approximately by 228.3% to 33.39 mm, by 45.9% to 21.73 mm, and by 14.9% to 15.77 mm, respectively. This region also experienced the generation of numerous new cracks. While in region B-4, visibly macroscopic original cracks extended by approximately 120% from 12.55 mm to 27.61 mm after testing, with a large area of new cracks being generated. Regions C-3 and C-4 suffered severe macroscopic damage, with well-developed new cracks and pore system after massive coal block collapse, significant original crack propagation, and increase of apertures.

Further, under high-pressure CO₂ conditions, original micro-cracks were not well-developed. But more new cracks were generated. Specifically, in regions A-1 and A-2, new cracks with lengths of 18.04 mm, 18.18 mm, and 29.98 mm were formed; in regions A-3 and A-4 new cracks with lengths of 26.44 mm and 47.06 mm were formed; in region C-2 new micro-cracks with lengths of 37.96 mm and 27.35 mm were formed after splitting damage, forming numerous new surfaces and pore systems in the upper region as coal particle collapse. In addition, numerous new surfaces based on original defects were generated in region C-3. Interconnected micro-cracks with lengths of approximately 64.11 mm were formed in region C-1, forming extensive new pore networks and micro-cracks with massive coal block collapse.

3.3. Quantitative Analysis of Coal Sample Damage After IDG-CF Tests

To more clearly compare the damage of coal samples before and after tests in quantity, the characteristic information of cracks was statistically extracted from Figure 8 and the difference in the three gas types was analyzed. The statistical results are shown in Figure 9. From the figures, it indicated that the vast majority of surface cracks in raw coal samples before tests under different gas conditions were micro-cracks with characteristic sizes of $d_{j,i}$ = 0–0.3 mm, accounting for approximately 90%. After testing, the absolute proportion of micro-cracks remained unchanged, but the distribution of macro-cracks changed to varying degrees.

As shown in Figure 9, it can be obviously observed that surface cracks and macro-defects of coal samples employed under CO₂ conditions were most developed before tests, which under N₂ conditions were the most intact, consistent with the characteristics in the physical photographs (see Figure 7 and Figure 8). This consistency further validated the rationality of the characteristic size analysis method used in this study, indicating that the characteristic size in Equation (2) could effectively describe the crack distribution features on coal sample surfaces and can be used for quantitative analysis of coal sample damage.

From Figure 9a, the proportion of micro-cracks on the coal sample surface increased significantly after N₂ tests, and newly generated micro-cracks with characteristic sizes of $d_{j,i}$ > 3 mm appeared, with the maximum characteristic size of $d_{j,i\max }$ = 10.05 mm. The average crack characteristic size $D_c$ before and after tests were 0.015 mm and 0.032 mm, respectively, with a change of $\Delta D_c$ = 0.017 mm. Figure 9b showed the same significant increase in micro-cracks proportion under CH₄ conditions, while the characteristic sizes of newly generated micro-cracks were $d_{j,i}$ > 4 mm, with the maximum value of $d_{j,i\max }$ = 10.75 mm. The average characteristic sizes $D_c$ before and after tests were 0.027 mm and 0.048 mm, respectively, with a change of $\Delta D_c$ = 0.021 mm. Under CO₂ conditions, as shown in Figure 9c, the micro-cracks proportion also increased and the new cracks’ characteristic sizes were $d_{j,i}$ > 4.5 mm, with the maximum value of $d_{j,i\max }$ = 17.1 mm. The average value $D_c$ before and after tests were 0.093 mm and 0.125 mm, respectively, with a change of $\Delta D_c$ = 0.032 mm.

In addition, the differences in distribution probability of crack characteristics sizes $\Delta P$ before and after tests under various gas conditions were compared, as shown in Figure 9d. The results demonstrated that the $\Delta P$ was increased and most pronounced under CO₂ conditions, with a maximum increment $\Delta P_{\max }$ reaching 1.40% and the largest value of maximum characteristic size $d_{j,i\max }$ of newly generated macro-cracks. Under CH₄ conditions, the maximum increment $\Delta P_{\max }$ was 0.51%, with the second largest value of $d_{j,i\max }$. Under N₂ conditions, the maximum increment $\Delta P_{\max }$ was 0.54%, with the smallest value of $d_{j,i\max }$. Although the maximum increment of crack characteristic size distribution probability $\Delta P_{\max }$ under CH₄ conditions was smaller than that under N₂ conditions, the probability increment $\Delta P$ of macro-cracks with characteristic sizes of ≥2 mm was significantly larger than that under N₂ conditions, the same as the maximum characteristic size $d_{j,i\max }$, which indicated that the damage extents of coal sample under CH₄ conditions was greater than that under N₂ conditions. Under CO₂ conditions, the probability increment $\Delta P$ of macro-cracks with characteristic sizes in the range of 0–2.4 mm was the largest. Despite the slightly smaller $\Delta P$ of macro-cracks with characteristic sizes ≥ 2.4 mm than that under CH₄ conditions, while its $d_{j,i\max }$ was the largest, indicating that the damage extents of coal sample tested with CO₂ was the greatest.

Furthermore, the mass, planar porosity, fractal dimensions of surface cracks, and the change rates of the coal samples before and after tests under different gas conditions were analyzed, as shown in Figure 10. From Figure 10a, the planar porosity increased by 2.29% with a growth rate of 128.2% after tests under N₂ conditions; increased by 2.63% with a growth rate of 155.4% under CH₄ conditions; and increased by 4.58% with a growth rate of 455.6% under CO₂ conditions. Figure 10b revealed that the fractal dimensions $D$ of crack networks on coal sample surfaces after tests increased with the enhancement of gas adsorption capacity, meaning the complexity of new crack networks also increased accordingly, further validating the intensification of sample damage extents. In addition, the outburst intensity $I$ detailed in Figure 10c, the relative change in coal sample mass after tests, showed the same increase trend as the gas adsorption capacity increasing. In conclusion, the trend of the aforementioned three parameters indicated that sample damage extents under different gas types followed the regularity of CO₂ > CH₄ > N₂. It is thus clear that the stronger the adsorption capacity of gas within coal mass, the more energy it carries during outbursts, and the higher the complexity of newly generated crack networks, the greater the coal mass damage extents, and the larger the outburst intensity.

3.4. Mechanical Response Characteristics of Coal Samples During IDG-CF Tests

The mechanical response characteristics of coal samples during IDG-CF tests under different gas conditions are shown in Figure 11. As illustrated in Figure 11a, upon the rupture of the bursting disc, the gas pressure within both the coal sample chamber and the coal matrix rapidly decreased to atmospheric pressure within 0.85 s. The rate of pressure decline was essentially consistent across the three tested gas types, indicating that the type of adsorbed gas has negligible influence on the gas pressure release rate. This phenomenon occurs because when the coal sample is instantaneously exposed to the atmosphere from a confined gas space, the free gas within the chamber and the pore structure of the coal is released first. While gas desorption within the coal sample theoretically initiates instantaneously, only gas desorbed from the external surface and the surfaces of open macropores [49]—which are immediately connected to the atmosphere upon depressurization—could contribute directly to the free gas release without requiring diffusion or seepage through the coal matrix.

Gas adsorption in coal predominantly occurs within complex microporous environments. Consequently, the quantity of gas desorbed and released from macropore surfaces is minimal relative to the total adsorbed gas content within the coal. Gas adsorbed on the internal pore surfaces must undergo diffusion and seepage processes for release, which occur over significantly longer timescales compared to instantaneous depressurization. Therefore, although the intrinsic desorption rates within the coal matrix differ depending on the gas type, the resulting variation in the total amount of gas released during the instantaneous depressurization is negligible under identical initial pressure conditions. This minute difference hardly exerts discernible impact on the overall gas release dynamics. Hence, the gas adsorption capacity exerts minimal influence on the gas pressure release rate. These findings demonstrate that the dynamic response of adsorbed gas desorption plays a pivotal role in the fragmentation of the coal sample.

A rarely reported axial stress evolution pattern was observed during tests, as depicted in Figure 11b. The results revealed that axial stress underwent a rapid decline immediately following the rupture of the bursting disc, followed by a swift recovery to nearly its initial level within an extremely short timeframe. This phenomenon was consistently observed across all tests, with the magnitude of stress fluctuation varying according to the type of adsorbed gas. Thus, this phenomenon cannot be attributed to free gas thin-layer pressure release or instrumental error, but represents a genuine physical response, a credible experimental phenomenon. Notably, a similar axial stress response was documented by Xu et al. [50] in their physical simulations of coal and gas outbursts.

The transient evolution of axial stress during instantaneous depressurization was characterized and quantified through Equation (4):

$$\begin{array}{l}\left\{\begin{array}{l}{\eta }_{\sigma d}={\displaystyle \frac{\Delta {\sigma }_d}{{\sigma }_{\max }}}\times 100\%,{\eta }_{\sigma u}={\displaystyle \frac{\Delta {\sigma }_d}{{\sigma }_{\min }}}\times 100\%\\ v_{\sigma }={\displaystyle \frac{\Delta {\sigma }_d}{\Delta t}}\end{array}\right.\\ \end{array}$$

(4)

where ${\eta }_{\sigma d}$ and ${\eta }_{\sigma u}$ are the stress decrease and rebound rate, respectively; $\Delta {\sigma }_d$ and $\Delta {\sigma }_u$ are the stress decrease and recovery magnitude, respectively; ${\sigma }_{\max }$ and ${\sigma }_{\min }$ are the maximum and minimum stress values, respectively; and $v_{\sigma }$ is the stress declining speed.

As shown in Figure 11b, when CO₂ was adsorbed, the stress decrease magnitude $\Delta {\sigma }_d$ was the largest, reaching 0.922 MPa, with a decrease rate ${\eta }_{\sigma d}$ of 16.28%. This was followed by CH₄, with $\Delta {\sigma }_d$ of 0.891 MPa and ${\eta }_{\sigma d}$ of 15.71%. In comparison, when the adsorbed gas was N₂, the stress decrease rate was the smallest, with $\Delta {\sigma }_d$ of 0.835 MPa and ${\eta }_{\sigma d}$ of 14.74%. After the sharp decrease, axial stress began to gradually increase, with speed and magnitude behaving with the same regularity as the decrease stage. The axial stress rebound rate ${\eta }_{\sigma u}$ when using N₂, CH₄, and CO₂ were 13.98%, 15.73%, and 17.22%, respectively. The axial stress decrease rate ${\eta }_{\sigma d}$, rebound rate ${\eta }_{\sigma u}$, and decrease speed $v_{\sigma }$ all increased with enhanced gas adsorption capacity, as illustrated in Figure 11d.

The deformation characteristics of coal samples during instantaneous depressurization under different gas conditions are revealed in Figure 11c. Upon the rupture of bursting disc, the rapid release of free-gas-induced radial expansion within the coal matrix occurs, triggering instantaneous bulk expansion that peaked the volumetric strain. As gas pressure continued to decline, the expansive effect weakened progressively. Concurrently, massive desorption of adsorbed gas contracted the coal matrix, driving overall sample contraction. This resulted in a gradual reduction in volumetric strain until stabilization at a residual value, attributable to irreversible deformation during structural failure. Notably, the magnitude of coal deformation exhibited a positive correlation with gas adsorption capacity (see Figure 11d). Enhanced adsorption capacity corresponded to greater volumetric strain, owing to the faster desorption kinetics of highly adsorbable gases liberating larger gas volumes within equivalent timeframes, carrying greater energy transfer during diffusion and seepage through pore structures, finally amplified expansion work. Consequently, gases with superior adsorption capacity induced more pronounced volumetric strain, further resulting in more severe damage of coal mass.

3.5. Influences of Gas Type on Outburst Coal-Breaking Under Gas–Stress Coupled Condition

Coal and gas outburst initiation requires both structural destabilization and mechanical failure of coal, primarily triggered by mining-induced stress redistribution. This phenomenon fundamentally links to in situ stress, gas pressure, tectonic features, and coal’s physical–mechanical characteristics [51]. After initial failure, residual energy progressively fragments coal, ejecting ultra-fine particulates (<0.075 mm) from the working face [52,53]. Current models attribute the energy sources to adsorbed/free gas expansion energy and coal matrix-stored elastic strain energy. Critically, overburden stress-derived elastic energy alone is insufficient to generate sub-75 μm coal particles, confirming gas’s essential role in achieving fragmentation during outbursts as an energy contributor. Specifically, gas accelerates coal fragmentation during outbursts through the dual pathways. Two main mechanisms are involved. First, adsorption–desorption dynamics induce coal matrix deformation via swelling/contraction effects, altering mechanical properties during the pre-outburst phase. Secondly, post-ejection expansion energy release contributes to outburst development through further pulverizing coal particles and mediating energy transfer. Consequently, gas ad/desorption kinetics constitute critical determinants for outburst initiation and sustainment. When adsorption equilibrium is disrupted, free-phase gas within coal pores is instantaneously released, followed by rapid conversion of adsorbed gas to the free state. Both free gas diffusion and adsorbed gas desorption release energy performing work on the coal matrix, termed gas expansion energy. The expansion work by gas during coal and gas outbursts was initially postulated as adiabatic [54,55], an assumption subsequently invalidated due to neglecting coal’s inherent properties as a porous geomaterial. The extensive gas–coal interfacial contact enables substantial convective heat transfer, even during transient outburst events, driving the process toward near-isothermal conditions via polytropic expansion. Consequently, the expansion energy for 1 mol of ideal gas is quantified by Equation (5).

$$W={\displaystyle \frac{RT}{n-1}}\left[1-{\left({\displaystyle \frac{p_0}{p}}\right)}^{{\displaystyle \frac{n-1}{n}}}\right]$$

(5)

where $P$ is the initial gas pressure; $P_0$ is the gas pressure after expansion; $n$ is the polytropic index, the ratio of gas constant pressure heat capacity $C_p$ to constant volume heat capacity $C_v$, $n=C_P/C_v$, taken as 1.0063; $R$ is the gas constant; and $T$ is the temperature of gas before expansion.

Thus, the work performed by the release and expansion of free gas in a unit mass of coal is

$$W_f={\displaystyle \frac{\phi }{\rho (n-1)}}\left[1-{\left({\displaystyle \frac{p_0}{p}}\right)}^{{\displaystyle \frac{n-1}{n}}}\right]p$$

(6)

where $\phi$ is the porosity of tested coal samples; $\rho$ is the apparent density of coal samples; $\phi$ = 0.0916, $\rho$ = 1.405 g/cm³ according to Table 1.

The adsorbed gas content in coal at specific pressures is typically calculated using the Langmuir isotherm model, which incorporates corrections for ash content and moisture content. To simplify computational procedures, temperature-dependent effects on adsorption capacity are omitted in this formulation. Thus, the work performed by the desorption of adsorbed gas is

$$W_a={\displaystyle \frac{\zeta RT}{V_m(n-1)}}\left[1-{\left({\displaystyle \frac{p_0}{p}}\right)}^{{\displaystyle \frac{n-1}{n}}}\right]\cdot {\displaystyle \frac{ap}{1/b+p}}$$

(7)

where $\zeta ={\displaystyle \frac{1}{1+0.31M}}\cdot {\displaystyle \frac{100-A-M}{100}}$, $A$ and $W$ are the ash content and the moisture content, respectively, taken according to Table 1; $V_m$ is the molar volume of gas, 22.4 L/mol for ideal gas; $a$ and $b$ are adsorption constants, taken according to Table 2.

Therefore, the expansion energy generated by the release of gas in the coal sample during instantaneous depressurization could be quantitated as

$$W_g=W_f+W_a$$

(8)

Following gas adsorption equilibrium (N₂, CH₄, CO₂) per unit coal mass, the expansion work during instantaneous depressurization could be quantified by Equation (8). It is evident from the equation that the gas expansion energy is proportional to the adsorption constant $a$ and $b$, indicating that enhanced adsorption performance directly increases expansion work capacity. Among gases tested in IDG-CF experiments, CO₂ exhibits the highest adsorption capacity, followed by CH₄, with N₂ showing the weakest adsorption. Upon equilibrium disruption, CO₂ desorbs most rapidly from coal matrices. Its sustained release transports greater energy, inducing more severe coal fragmentation during instantaneous depressurization. Furthermore, prior studies confirm that adsorbed gases deteriorate coal mechanical properties to varying extents, with degradation severity scaling proportionally to gas adsorption capacity [23]. During the IDG-CF tests, adsorbed gases significantly compromised coal sample integrity, resulting in quantifiable mechanical degradation. Consequently, coal fragmentation intensity during outbursts depends critically on both the adsorption performance of in situ gases and the gas-induced deterioration of coal mechanical properties.

Therefore, for mining gas-bearing coal seams—particularly outburst-prone seams—the gas occurrence state within coal significantly influences outburst risks during extraction. Stronger gas adsorption capacity leads to higher efficiency of desorption from adsorbed to free states, faster diffusion rates of free-state gas, and greater energy release during state transition. Consequently, gas expansion performs more work on coal, causing severe damage. Additionally, gases with stronger adsorption exhibit a more pronounced weakening effect on coal’s mechanical properties, resulting in a drastic reduction in mechanical strength and diminished resistance to failure. Thus, coal seams containing highly adsorbent gases (e.g., CO₂ > CH₄ > N₂) face higher outburst probabilities during mining. This underscores the necessity of proactively exploring gas occurrence states and implementing effective gas pre-drainage techniques before mining, ensuring both safe coal production and national energy security.

4. Conclusions

This study simulated the coal-breaking process of gas–stress coupled during outburst occurrence through instantaneously depressurized gas-induced coal fragmentation tests, and conducted in-depth analysis of coal sample damage extents and deformation characteristics under different gas conditions. The main conclusions are as follows:

• The image binarization method was creatively used to quantify the coal sample damage before and after IDG-CF tests. It indicated that the stronger adsorption capacity of gas within coal mass resulted in more complex newly generated crack networks, greater coal damage extents, and larger outburst intensity.
• A rarely reported axial stress rebound phenomenon was observed in tests. The mechanical response characteristic analysis results proved that the dynamic response during adsorbed gas desorption played a key role in coal sample damage.
• From the perspective of expansion energy, the stronger the adsorption capacity of gas within coal seams, the more energy gas carries and the more pronounced coal weakening, inducing higher outburst probabilities.

In essence, enhanced gas adsorption capacity accelerates desorption kinetics and energy release, intensifying coal damage during gas–stress coupled instantaneous depressurization. These results advance the mechanistic understanding of outburst dynamics and provide theoretical foundations for disaster prevention and early warning systems. The findings could also potentially contribute to the development of more precise outburst prediction models and real-time monitoring system and may offer practical insights for optimizing deep coal seam mining operations and gas extraction strategies, ultimately supporting global energy security. However, this study focuses on experimental research without involving numerical simulation analysis. In future research, we would conduct numerical simulation studies to further improve theoretical understanding of coal and gas outburst mechanisms.