Experimental Study on the Mechanical Properties and Acoustic Emission Characteristics of Deep Soft Rocks under Low-Frequency Dynamic Disturbance

Abstract

The strong dynamic disturbance in deep mines seriously affects the safe and efficient mining of deep resources. In this study, we used the creep disturbance impact loading system and acoustic emission system to conduct experiments on soft siltstone specimens under a combination of dynamic and static loads. Based on the failure characteristics and waveform signals, the mechanical properties and acoustic emission characteristics of soft rocks under different dynamic disturbances were quantitatively revealed. The experimental results show that: (1) Under the dynamic disturbance, the deformation of the siltstone specimens increases as the initial average stress increases. When the axial stress exceeds the upper stress threshold, cracks continue to propagate, resulting in the destabilization of the specimen. (2) The magnitude of the initial average stress is closely related to the degree of damage and failure mode of the siltstone. With the increase in the initial average stress, the failure mode of the siltstone specimens gradually changes. As the initial average stress increases, the maximum load first decreases, then increases, and finally decreases, and the fitted curve is polynomial. We used the RFPA2D cyclic loading module to analyze the effect of the elastic modulus of each loading step on the damage evolution of the specimen under dynamic disturbance. The waveform characteristics during the evolution of the damage of the specimens were analyzed by extracting signals at the key points.

1. Introduction

Of the current energy and mineral reserves, coal is a strategic energy source supporting China’s economic and social development [1,2]. At present, deep mining has become the new normal for coal resource development in China. Safety issues during deep mining and major disaster prevention and control have been written into China’s “14th Five-Year Plan” [2]. Under complex geological conditions, deep mines are strongly disturbed by the mining dynamics; therefore, large deformation hazards and ground control in deep soft rock mines have become the focus of research in the field of deep mining [1,3,4]. The study of mechanical properties of soft rocks at depth under low-frequency dynamic disturbance is of important practical significance for understanding the deformation destabilization mechanism and stability control of soft rock mines [5,6].

Regarding the dynamic disturbance to coal and rocks at depth, scholars have conducted extensive studies on experimental mechanical properties and acoustic emission characteristics of rocks under dynamic and static loading conditions.

Since 2012, dozens of rock bursts or suspected rock bursts have occurred in the Asher copper mine in Xinjiang, with the most serious one occurring during the anchor network spray concrete support, where the surrounding rock was flaky and ejected over a large area for a short period of time [7]; the Chener gold mine in Shaanxi explored at a depth of about 1000 m, and rock bursts occurred from time to time after entering deep mining, resulting in flakes falling off the top of the roadway and the two gangs, and fragments rapidly flying away [8]. By studying mechanical properties under combined dynamic and static loads, experiments were performed using the split Hopkinson pressure bar (SHPB) to examine the damage characteristics and the energy dissipation behavior of strain-type rock bursts at depth [9,10,11]. Mechanical experiments under different dynamic perturbations were carried out to investigate the energy-based failure mechanism of rocks, the evolution of fracture characteristics, and the effect of roof dynamic disturbance on the stress of the floor [12,13,14,15]. By studying rock bursts near a deeply buried tunnel fault under static and dynamic loads, static and dynamic loads were used to simulate rock damage. A combined finite element method was used to simulate the damage process of an underground cavern, revealing the damage mechanism of deep hard rock influenced by the amplitude of dynamic stress wave, disturbance direction and dip angle of the structural surface [16,17]. Dynamic disturbance, and cyclic loading is related to the fatigue crack expansion of the material, and when the capacity of the material is exceeded, the material undergoes damage [18,19,20].

In studying acoustic emission (AE) characteristics of coal and rocks at depth, the rock failure process analysis (RFPA) code was used to analyze the influencing factors on AE characteristics of rock-like specimens at different positions, and the evolution of permeability of fractured rock and its anisotropy under hydrodynamic coupling conditions [21,22]. Zhang et al. [23] analyzed the evolution of acoustic emission and deformation field during crack propagation by establishing a mechanistic model for directional crack propagation. Researchers have carried out shear experiments on granite and triaxial compression experiments on coal specimens of different sizes to explore the variations of transverse and longitudinal AE wave signals and the evolution of energy [24,25,26,27,28]. Acoustic emission signals of rocks during damage and the influence of different fracture dip angles on the macroscopic mechanical properties of rocks were investigated by using uniaxial compression experiments with AE monitoring [29,30]. The torsional properties and damage development of long and hollow axisymmetric composite cylinders with different opening sizes were investigated experimentally. The acoustic emission (AE) of prismatic marble specimens during uniaxial compression loading was analyzed using the non-extensive statistical mechanics (NESM) method. Acoustic emission (AE) was used to record the microcracks during the impact of granite specimens [31,32,33]. Different materials can be studied using acoustic emission and digital image correlation techniques to investigate the specific characteristics of the material damage process under different loading effects in order to follow the evolution of the material under different damage modes and to determine the strain [34,35,36,37].

In summary, most of these experimental studies on the mechanical properties of coal and rock under combined dynamic and static loads were focused on single, high-frequency impact loading using SHPB and drop hammers. By applying the average value of different cyclic loading stress values, the siltstone is damaged under the action of dynamic perturbation, and the deformation damage characteristics of the dynamic perturbation are necessarily different from those of uniaxial. To this end, we select the siltstone from the roof of a deep soft rock tunnel in central eastern China as the research object and use the creep disturbance dynamic impact loading system combined with acoustic emission monitoring to carry out experiments under different dynamic disturbance conditions and reveal the mechanical and acoustic emission characteristics of the soft rock. Our study aims to provide a theoretical basis for the control of the deformation and instability of coal and rocks in deep mines affected by mining activities.

The study of the effect of low-frequency dynamic disturbances on the mechanical properties of rocks is of greater engineering interest. In rock mechanics tests, the dynamic disturbance environment in the downhole mining process can be simulated by applying a certain frequency of cycles to the rock specimen. The magnitude of stress perturbation varies depending on the form of dynamic perturbation such as blasting and stress redistribution. In deep well mining, the rock is in a high stress state and a small stress perturbation may cause the rock to exceed its yield stress and cause damage to the rock. In summary, most of these experimental studies on the mechanical properties of coal rocks under combined dynamic and static loads have focused on single high-frequency impact loading with SHPB and drop hammers. By applying average values of stress values for different cycles of loading, the siltstone was damaged by dynamic disturbance, which inevitably has different deformation damage characteristics from that of uniaxial ones. To this end, we selected a siltstone in the roof of a deep soft rock tunnel in east-central China as the research object and used a creep perturbation dynamic impact loading system combined with acoustic emission monitoring to conduct experiments under different dynamic perturbation conditions to reveal the mechanical characteristics of the soft rock. Our study aims to provide a theoretical basis for controlling the deformation and instability of coal and rocks in deep mines affected by mining activities.

2. Experimental Scheme and Conditions

2.1. Experimental System

Experiments are conducted by using a creep disturbance dynamic impact loading system, which mainly consists of a loading system, a hydraulic pump station, a servo control system, and a data acquisition system. The maximum axial load of the static loading unit in the loading system is 800 kN, and the maximum axial load of the dynamic loading unit is 100 kN, as shown in Figure 1. The servo control and data acquisition system consists of a high-precision closed-loop servo control unit, pressure transducer, and magneto-displacement transducer to form a closed-loop control system to achieve accurate monitoring and acquisition of stress and displacement in a variety of stress paths. The acquisition system mainly includes indenter displacement and axial stress data acquisition, and servo control through the supporting software. Axial stress is monitored by pressure sensors and recorded with real-time feedback. Axial deformation and axial stress of the specimen are automatically collected, and acoustic emission monitoring is used to track the crack damage process. Multiple sets of cyclic loading experiments are applied using sine wave loading.

2.2. Specimen Preparation

The siltstone used in this work was recovered from the roof of a mining roadway in Xinshanghai NO.1 Coal Mine in China, which is a typical Jurassic soft rock coal mine. The samples were brownish-grey, with no visible structural defects on the surface. Specimen preparation procedure: (1), samples were taken from the same rock block to make standard cylindrical specimens of Φ50 mm × 100 mm. (2), the two ends of the specimens were polished by using a stone grinding machine so that the non-parallelism and non-perpendicularity were less than 0.2 mm, in accordance with the requirements of the International Society of Rock Mechanics (ISRM) and (3) finally, the specimens were divided into two groups, numbered RS-1~3 and RS-4~24, with a natural apparent density of 2325.4~2630.5 kg/m³ and an average density of 2477.95 kg/m³, as shown in Figure 2.

2.3. Experimental Design

The siltstone was divided into two groups for uniaxial compression experiments and cyclic disturbance loading experiments. During the initial compression loading stage, displacement-controlled loading (4 μm/s) was applied to compact the microfractures inside the specimens. In the cyclic loading stage, specimens were first loaded to 5.48 MPa, 6.39 MPa, 7.31 MPa, 9.13 MPa, 10.05 MPa, 10.96 MPa, and 11.87 MPa (30%, 35%, 40%, 50%, 55%, 60%, and 65% of the mean uniaxial compressive strength). The sine wave cyclic loading method was then used to apply the dynamic loads. The uniaxial compressive strength of 18.264 MPa, obtained from the uniaxial compression tests, was used as the initial average of the cyclic load. The experimental sinusoidal loading waveform is shown in Figure 3. The specimen loading process was as follows: first, the specimen was loaded at a rate of 0.1 mm/min to the initial value of the cyclic stage, and then the sinusoidal waveform was cycled 1000 times until the siltstone specimen was damaged or the axial force was 0. A schematic diagram of the cyclic disturbance loading for the siltstone specimens is shown in Figure 3. By analyzing the relationship between different initial average stresses and axial strain and elastic modulus, the mechanical properties of siltstone under sinusoidal cyclic loading can be better understood, and the fitting trend of the maximum load with different initial average stresses can be seen more graphically.

3. Analysis of Experimental Results

3.1. Uniaxial Compression Process and Experimental Results

The experimental design and physical parameters of the siltstone are provided in

Table 1. Testing program.

Specimen Number	Quality/g	Height/mm	Diameter/mm	Density/(kg·m−3)	Initial Average Stress Value/MPa	Compressive Strength/MPa
RS-1	467.49	100.18	50.20	2358.9	—	17.74
RS-2	452.52	99.86	49.53	2353.1	—	19.27
RS-3	464.57	100.16	50.07	2356.9	—	17.78
RS-4	458.63	100.02	50.12	2325.4	5.48	15.29
RS-5	467.58	100.11	50.08	2372.3	5.48	13.47
RS-6	463.86	100.13	50.01	2359.7	5.48	14.25
RS-7	460.28	100.01	50.09	2336.7	6.39	13.40
RS-8	462.49	100.20	50.03	2349.1	6.39	13.71
RS-9	464.82	100.18	50.05	2359.5	6.39	12.81
RS-10	465.25	100.28	50.05	2359.4	7.31	14.38
RS-11	481.52	100.15	50.12	2438.2	7.31	13.71
RS-12	470.29	100.18	50.13	2379.6	7.31	12.92
RS-13	466.76	99.98	50.03	2376.0	9.13	16.47
RS-14	462.09	100.20	50.12	2338.6	9.13	13.35
RS-15	463.73	100.21	50.03	2355.2	9.13	15.39
RS-16	482.86	100.35	50.03	2449.0	10.05	13.31
RS-17	474.18	100.02	50.06	2409.9	10.05	16.08
RS-18	475.16	100.10	50.12	2407.2	10.05	13.45
RS-19	471.15	100.21	50.11	2385.2	10.96	13.71
RS-20	460.00	100.23	50.10	2329.2	10.96	13.64
RS-21	520.18	100.32	50.11	2630.5	10.96	13.57
RS-22	467.45	100.14	50.02	2376.7	11.87	8.83
RS-23	484.84	100.10	50.01	2467.1	11.87	9.85
RS-24	473.13	100.08	50.10	2399.4	11.87	9.42

, where RS-1, RS-2 and RS-3 are uniaxial compression tests and the rest are cyclic disturbance loading tests. The stress-strain curves of uniaxial compression tests and the comparison of the specimens pre- and post-failure are shown in Figure 4, where the peak stress in the stress-strain curve is the compressive strength. The uniaxial compression stress-strain curve exhibits four stages: microfracture compaction stage (I), elastic deformation stage (II), yielding stage (III), and post-peak stage (IV).

It can be seen from Figure 4 that during uniaxial compression, elastic modulus increases first and then decreases as the axial stress increases, indicating that when the axial stress is applied at the beginning of the experiment, the cracks inside the specimen are compacted, leading to an increase in elastic modulus. When the axial stress is applied to a certain extent, the cracks inside the specimen begin to propagate and then the main cracks coalesce, leading to a decrease in elastic modulus. Afterward, the axial stress continues to increase until it exceeds the bearing capacity of the specimen and then the specimen is ruptured.

3.2. Effect of Dynamic Disturbance on the Failure Pattern of Siltstone Specimens

Figure 5 shows the stress-strain curves of the specimens under various dynamic disturbances. When the initial average stresses of the specimens are similar, the stress-strain curves are also similar. However, there are differences in the stress-strain curves when the initial average stresses are different. It can be seen from the figure that the fracture compaction occurs mostly at the beginning of the loading, and the curve is upward convex in the compaction stage. After that, when the sine wave cyclic disturbance load σ_c for the siltstone specimens is greater than 60% of the uniaxial compressive strength, the specimens could be failed before the preset number of cycles (1000) is completed. During the cyclic dynamic disturbance loading, the hysteresis loop area indicating the energy loss of the specimen during the cyclic loading changes with σ_c. When the initial average stress value is 5.48 MPa (30% of σ_c), the hysteresis loop is “thin-lobed”, and when the initial average stress value is 6.39 MPa (35% of σ_c), the hysteresis loop becomes “full” and its area increases. This phenomenon occurs due to the large deformation of the primary fractures in the specimen at the beginning of the cyclic dynamic disturbance stage. The hysteresis loops become “sharp-lobed” when the initial cyclic load is 7.31 MPa, 9.13 MPa (i.e., 40% and 50% of σ_c). When the initial cyclic load is 10.05 MPa (55% of σ_c), the hysteresis loop is “banded” and its area gradually decreases, indicating that the fractures inside the specimen are compacted and closed, and the energy dissipation is reduced. When the average stress is 10.96 MPa (60% of σ_c), the hysteresis loop becomes “full” and the specimen undergoes greater irreversible deformation, resulting in unstable failure. When the average stress is 11.87 MPa (65% of σ_c), the specimen is failed before the cycle starts and there is no hysteresis loop.

From the axial stress-strain curves of these seven specimens, it can be found that the deformation of the siltstone specimens increases as the initial average stress increases, indicating that the damage of the specimens during the cyclic disturbance also increases. When the axial stress exceeds the upper limit threshold of stress, the cracks continue to extend and intersect with each other, which in turn leads to the destabilization of the specimens.

Figure 6 shows a typical specimen failure plot under different initial average stresses. When the initial average stress value is 5.48 MPa, its impact on the damage of the specimen is low, and its damage characteristics are generally similar to that from uniaxial compression tests. Due to the low initial average stress, the specimen does not show obvious macroscopic damage after 1000 cycles of disturbance, and only new microfractures are created. The failure mode is shear, and there is partial damage in the middle of the specimen. When the initial average stress is 6.39 MPa, due to axial compressive stress, the shear stress on the fracture surface exceeds the shear strength and the tensile stress in the radial direction. The failure mode is mainly a shear failure, and there is a small part of splitting failure, and the specimen becomes cone-shaped inside after the failure. When the initial average stress is 7.31 MPa, the axial compressive stress causes the radial tensile stress of the specimen; once the radial tensile stress exceeds the tensile strength of the rock, split failure occurs; at this time, the extension direction of secondary cracks changes from longitudinal to dendritic dispersion. The initial average stress of 9.13 MPa has a greater impact on the overall failure mode of the specimen: a large number of secondary cracks are created in the specimen, and large pieces of rock fragments are peeled from the surface, and the failure mode is a combination of shear and splitting. When the initial average stress is 10.05 MPa, the failure mode is shear, with a clearly visible vertical main crack penetrating through the specimen. When the initial average stress is 10.96 MPa, the macroscopic crack of the specimen becomes the shear surface inclined at 60 degrees from the top to the bottom. When the initial average stress is 11.87 MPa, the specimen can no longer bear axial stress and cyclic disturbance, the failure mode is tensile/shear mixed, the integrity of the failed specimen is poor, the rupture surface is rough, showing the characteristics of brittle failure. When the specimen is ruptured, rock fragments burst with a loud sound.

Therefore, it can be seen from the macroscopic failure of the specimen that under the cyclic dynamic disturbance, the failure mode gradually transitions from shear failure to shear-splitting mixed failure. Under large initial average stress (i.e., 18.264 MPa), a higher degree of scattering and fragmentation is observed, and the main crack extends to the bottom of the specimen, which shows that the failure mode of the specimen is closely related to the magnitude of the initial average stress.

3.3. Numerical Model Development and Validation

The RFPA2D cyclic loading module was used for the numerical simulation. The RFPA software used in this paper, Rock Failure Process Analysis System (RFPA), is a tool based on an elastic damage model that takes into account the non-uniformity of rock materials and the distribution of defects using the intrinsic relationship of statistical damage randomness of defect distribution. We first established a two-dimensional plane strain model with a length × height of 50 mm × 100 mm with 50 elements in the X direction and 100 elements in the Y direction. The homogeneity of the model is 3, the density is 2477.95 kg/m³, the damage criterion is the Mohr-ulomb criterion, the number of cyclic loading is 1000 times, the Poisson’s ratio is 0.25, and the uniaxial compressive strength is 18.264 MPa. The model is numerically modeled according to the corresponding dimensions, and Figure 7 shows a cyclic disturbance model based on the Mohr-Coulomb failure criterion. In the sample model, a pre-pressure is applied to the upper boundary, followed by a sinusoidal cyclic loading with a setting of 1000 times, and no force is applied to the lower boundary. Taking the initial average stress of 10.96 MPa as an example, the elastic modulus of each loading step is shown in Figure 8, which can reproduce the macroscopic crack development of the specimen. The evolution of acoustic emission signals is shown in Figure 9.

As can be seen in Figure 8, at the beginning of the cyclic disturbance loading, the specimen is relatively intact, and there are no noticeable cracks on the surface. As the axial stress increases, shear cracks are created with an inclination of 60 degrees, which gradually penetrate through the specimen. The acoustic emission events of the siltstone specimen during the damage evolution are shown in Figure 9. The center of each circle in the figure represents the location of the acoustic emission event, and the size of the circle represents the amount of energy released by the event. Each circle represents an acoustic emission event, where a white circle represents an acoustic emission event resulting from compression-shear failure, and a red circle represents an acoustic emission event from tensile failure. At the beginning of the cyclic disturbance loading, sporadic acoustic emission events occur locally in the model, and with the increase in axial stress, acoustic emission events are concentrated, gradually forming macroscopic fractures, which in turn leads to instability failure of the siltstone specimen.

3.4. Elastic Modulus Strengthening Effect at Different Initial Average Stresses

Figure 10 shows the number of cycles of siltstone specimens under different initial average stresses versus axial strain. The relationship between the cumulative axial strain and cyclic number can be divided into two cases: one is that when the initial average stress is small, no significant damage occurs in the siltstone specimen after completing 1000 cyclic disturbances and the other is that when the initial average stress is large, the siltstone specimen is damaged before completing 1000 cyclic disturbances or before the cycle begins. When the initial average stress was increased from 5.48 MPa to 10.05 MPa, the axial strain of the siltstone specimen increased slightly under cyclic disturbance, and then remained basically constant. When the initial average stress was 10.96 MPa, the axial strain increased with the increase in the number of cycles, and destabilization damage occurred at 219 cycles. When the initial average stress was 11.87 MPa, the specimen was damaged before the cycle started. In short, when the initial average stress does not reach a specific value, the specimen would not be damaged even if it is cyclically loaded multiple times, which indicates that there is an upper threshold for the damage of the siltstone specimen during cyclic loading.

Figure 11 shows the relationship between the elastic modulus of the siltstone specimen and the different initial average stresses, where A, B, C, D, E, Fand G represent the initial average stresses of 5.48 MPa, 6.39 MPa, 7.31 MPa, 9.13 MPa, 10.05 MPa, 10.96 MPa and 11.87 MPa, respectively. Taking the specimen with an initial average stress of 5.48 MPa as the reference, when the initial average stress was 6.39 MPa, 7.31 MPa, 9.13 MPa, 10.05 MPa, 10.96 MPa, and 11.87 MPa, the elastic modulus of the specimen increased by 15.5%, 19.6%, 49%, 59.2%, 64.5%, and 72.7%, respectively. The average of the elastic modulus of each group is used for data fitting, which shows that the elastic modulus increases as the initial average stress increases. Under the cyclic disturbance load, the pre-existing cracks in the siltstone specimen are closed, resulting in an increase in the stiffness and elastic modulus of the specimen.

The percentages corresponding to the different initial average stresses of the siltstone specimens are 30%, 35%, 40%, 50%, 55%, 60%, and 65%, and their corresponding maximum load relationship curves are shown in Figure 12. The testing process of the specimens can be divided into three stages, and the maximum load shows a trend of first decrease, then increase, and finally decrease, and the fitting curves are polynomial.

3.5. Characterization of Acoustic Emission Waveforms

The acoustic emission system can monitor the elastic waves during the creation of fractures inside the specimen and the corresponding energy change during cyclic disturbance loading. The key points N1 and N2 are extracted according to the acoustic emission energy release characteristics. N1 is the initial period of specimen dynamic disturbance loading, and N2 is when the damage to the specimen occurs. The acoustic emission characteristics at key points N1 and N2 are extracted and the signals are transformed from the time domain to the frequency domain by using the fast Fourier transform to obtain the principal frequency eigenvalues. MATLAB is used to obtain a three-dimensional plot of amplitude-frequency-time, and then the effect of the dynamic disturbance on the acoustic emission characteristics of the siltstone is analyzed.

Figure 13 shows the three-dimensional plot of the principal frequency characteristics obtained by the Fourier transform of the acoustic emission key points N1 and N2 of the specimens under different stresses, and the sharp points corresponding to the amplitude are the principal frequency eigenvalues. In different sets of cyclic disturbance experiments, the specimen changes from the high principal frequency at key point N1 to the low principal frequency at key point N2, and the principal frequency eigenvalues are reduced to a large extent. The high-frequency eigenvalues of the waves indicate small-scale damage to the specimen, and the low-frequency eigenvalues of the waves indicate large-scale damage to the specimen. The principal frequency eigenvalues of key point N1 ranges from 172 KHz to 203 KHz. The initial average stresses are 10.96 MPa and 10.05 MPa, corresponding to the lowest and highest principal frequency eigenvalues, respectively. The principal frequency eigenvalues of key point N2 ranges from 73 KHz to 98 KHz. The initial average stresses are 9.13 MPa and 6.39 MPa, corresponding to the lowest and highest principal frequency eigenvalues, respectively. The principal frequency values of key points of the siltstone specimens under different average stresses are shown in

Table 2. Principal frequency values of key points in siltstone specimens at different initial average stresses.

Key Point	5.48 MPa	6.39 MPa	7.31 MPa	9.13 MPa	10.05 MPa	10.96 MPa	11.87 MPa
N1	201 KHz	185 KHz	181 KHz	179 KHz	203 KHz	172 KHz	178 KHz
N2	87 KHz	98 KHz	85 KHz	73 KHz	82 KHz	79 KHz	87 KHz

.

4. Conclusions

(1)

The deformation of the siltstone specimens increases as the initial average stress increases, indicating that the damage of the specimen during the cyclic disturbance process also increases. When the axial stress exceeds the threshold of the upper stress limit, the cracks continue to propagate and coalesce, which in turn leads to the destabilization of the specimen.

(2)

The magnitude of the initial average stress is closely related to the degree of damage and failure mode of the siltstone. When the initial average stress is low, the failure model is mainly shear failure. With the increase in the initial average stress, the failure mode of the siltstone specimen gradually transitions from shear failure to shear-splitting mixed failure. The larger the initial average stress, the higher the degree of fragmentation of the specimen. The primary crack is developed from the direction of the maximum principle stress to the bottom of the specimen.

(3)

In the RFPA2D cyclic loading simulation, when the dynamic load starts, sporadic acoustic emission events are generated locally in the model. With the increase in axial stress, acoustic emission events become localized and macroscopic fractures are gradually formed, which in turn, leads to the destabilization of the siltstone specimen.

(4)

As the initial average stress increases, the modulus of elasticity increases. The reason is that under the dynamic disturbance loading, the pre-existing cracks within the siltstone specimens are closed, which increases the stiffness and elastic modulus of the specimen.